HANDBOOK OF OPERATIONS RESEARCH IN NATURAL RESOURCES pdf

Hillier, Series Editor, Stanford University Gass & Assad/ AN ANNOTATED TIMELINE OF OPERATIONS RESEARCH: An Informal History Greenberg/ TUTORIALS ON EMERGING METHODOLOGIES AND APPLICAT

HANDBOOK OF OPERATIONS RESEARCH

IN NATURAL RESOURCES

OPERATIONS RESEARCH & MANAGEMENT SCIENCE

Frederick S Hillier, Series Editor, Stanford University

Gass & Assad/ AN ANNOTATED TIMELINE OF OPERATIONS RESEARCH: An Informal History

Greenberg/ TUTORIALS ON EMERGING METHODOLOGIES AND APPLICATIONS IN

OPERATIONS RESEARCH

Weber/ UNCERTAINTY IN THE ELECTRIC POWER INDUSTRY: Methods and Models for Decision

Support

Figueira, Greco & Ehrgott/ MULTIPLE CRITERIA DECISION ANALYSIS: State of the Art Surveys

Reveliotis/ REAL-TIME MANAGEMENT OF RESOURCE ALLOCATIONS SYSTEMS: A Discrete Event

Systems Approach

Kall & Mayer/ STOCHASTIC LINEAR PROGRAMMING: Models, Theory, and Computation

Sethi, Yan & Zhang/ INVENTORY AND SUPPLY CHAIN MANAGEMENT WITH FORECAST

UPDATES

Cox/ QUANTITATIVE HEALTH RISK ANALYSIS METHODS: Modeling the Human Health Impacts of

Antibiotics Used in Food Animals

Ching & Ng/ MARKOV CHAINS: Models, Algorithms and Applications

Li & Sun/ NONLINEAR INTEGER PROGRAMMING

Kaliszewski/ SOFT COMPUTING FOR COMPLEX MULTIPLE CRITERIA DECISION MAKING

Bouyssou et al/ EVALUATION AND DECISION MODELS WITH MULTIPLE CRITERIA: Stepping

stones for the analyst

Blecker & Friedrich/ MASS CUSTOMIZATION: Challenges and Solutions

Appa, Pitsoulis & Williams/ HANDBOOK ON MODELLING FOR DISCRETE OPTIMIZATION

Herrmann/ HANDBOOK OF PRODUCTION SCHEDULING

Axsäter/ INVENTORY CONTROL, 2 nd Ed

Hall/ PATIENT FLOW: Reducing Delay in Healthcare Delivery

Józefowska & Węglarz/ PERSPECTIVES IN MODERN PROJECT SCHEDULING

Tian & Zhang/ VACATION QUEUEING MODELS: Theory and Applications

Yan, Yin & Zhang/ STOCHASTIC PROCESSES, OPTIMIZATION, AND CONTROL THEORY

APPLICATIONS IN FINANCIAL ENGINEERING, QUEUEING NETWORKS, AND MANUFACTURING SYSTEMS

Saaty & Vargas/ DECISION MAKING WITH THE ANALYTIC NETWORK PROCESS: Economic,

Political, Social & Technological Applications w Benefits, Opportunities, Costs & Risks

Yu/ TECHNOLOGY PORTFOLIO PLANNING AND MANAGEMENT: Practical Concepts and Tools

Kandiller/ PRINCIPLES OF MATHEMATICS IN OPERATIONS RESEARCH

Lee & Lee/ BUILDING SUPPLY CHAIN EXCELLENCE IN EMERGING ECONOMIES

* A list of the early publications in the series is at the end of the book *

HANDBOOK OF OPERATIONS RESEARCH

Diego Portales University, Santiago, Chile

University of Chile, Santiago, Chile

with the collaboration of

University of Chile Technical University of Madrid

University of Chile Portsmouth, United Kingdom Santiago, Chile

Fred Hillier

Stanford University

Stanford, CA, USA

Library of Congress Control Number: 2007924350

Printed on acid-free paper

© 2007 by Springer Science+Business Media, LLC

All rights reserved This work may not be translated or copied in whole or in part without the written permission of the publisher (Springer Science+Business Media, LLC, 233 Spring Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now know or hereafter developed is forbidden The use in this publication of trade names, trademarks, service marks and similar terms, even if the are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights

CEMARE, University of Portsmouth

Contributing Authors ixPreface xiiiAcknowledgments xv

Importance of Whole-Farm Risk Management in Agriculture 3

RUUD HUIRNE,MIRANDA MEUWISSEN,

AND MARCEL VAN ASSELDONK

Dealing with Multiple Objectives in Agriculture 17

KIYOTADA HAYASHI

Modeling Multifunctional Agroforestry Systems with Environmental

Values: Dehesa in Spain and Woodland Ranches in California 33

PABLO CAMPOS,ALEJANDRO CAPARRÓS,EMILIO CERDÁ,

LYNN HUNTSINGER, AND RICHARD B.STANDIFORD

Environmental Criteria in Pig Diet Formulation with Multi-Objective

TERESA PEÑA,CARMEN CASTRODEZA , AND PABLO LARA

Modeling the Interactions Between Agriculture and the Environment 69

SLIM ZEKRI AND HOUCINE BOUGHANMI

MCDM Farm System Analysis for Public Management of Irrigated

Agriculture 93

JOSÉ A.GÓMEZ-LIMÓN,JULIO BERBEL, AND MANUEL ARRIAZA

Water Public Agencies Agreeing to A Covenant for Water Transfers:

ENRIQUE BALLESTERO

Positive Mathematical Programming for Agricultural and

Environmental Policy Analysis: Review and Practice 129

BRUNO HENRY DE FRAHAN,JEROEN BUYSSE,PHILIPPE POLOMÉ,

BRUNO FERNAGUT,OLIVIER HARMIGNIE,LUDWIG LAUWERS,

GUIDO VAN HUYLENBROECK, AND JEF VAN MEENSEL

RAGNAR ARNASON

TROND BJØRNDAL AND GORDON MUNRO

VEIJO KAITALA AND MARKO LINDROOS

LINDA NØSTBAKKEN AND JON M.CONRAD

DANIEL E.LANE

Capacity and Technical Efficiency Estimation in Fisheries:

Parametric and Non-Parametric Techniques 273

SEAN PASCOE AND DIANA TINGLEY

Studies in the Demand Structure for Fish and Seafood Products 295

FRANK ASCHE,TROND BJØRNDAL, AND DANIEL V.GORDON

RAFAEL EPSTEIN,JENNY KARLSSON,MIKAEL RÖNNQVIST,

AND ANDRES WEINTRAUB

Log Merchandizing Model Used in Mechanical Harvesting 378

HAMISH MARSHALL

RAFAEL EPSTEIN,MIKAEL RÖNNQVIST, AND ANDRES WEINTRAUB

JOHN HOF AND ROBERT HAIGHT

ALAN T.MURRAY

JOHN SESSIONS,PETE BETTINGER, AND GLEN MURPHY

Forestry Economics: Historical Background and Current Issues 449

RONALD RAUNIKAR AND JOSEPH BUONGIORNO

Multiple Criteria Decision-Making in Forest Planning:

LUIS DIAZ-BALTEIRO AND CARLOS ROMERO

DAVID L.MARTELL

A Model for the Space–Time Spread of Pine Shoot Moth 511

ROBERTO COMINETTI AND JAIME SAN MARTÍN

Adaptive Optimization of Forest Management in A Stochastic World 525

PETER LOHMANDER

MINING 545

Application of Optimisation Techniques in Open Pit Mining 547

LOUIS CACCETTA

CHRISTOPHER ALFORD,MARCUS BRAZIL, AND DAVID H.LEE

Long- and Short-Term Production Scheduling at Lkab’s Kiruna Mine 579

ALEXANDRA M.NEWMAN,MARK KUCHTA,

AND MICHAEL MARTINEZ

An Integrated Approach to the Long-Term Planning Process

RODRIGO CARO,RAFAEL EPSTEIN,PABLO SANTIBAÑEZ,

AND ANDRES WEINTRAUB

Manuel Arriaza

Andalusian Institute of Agricultural,

Fisheries and Food Research

Spanish Council for Scientific

Research (CSIC), Spain

acaparros@ieg.csic.es

José A Gómez-Limón

University of Valladolid limon@iaf.uva.es

Bruno Henry de Frahan

Université Catholique de Louvain henrydefrahan@ecru.ucl.ac.be

University of Calgary

North Central Research Station dgordon@ucalgary.ca

Mikael Rönnqvist

The Forestry Research Institute

of Sweden

mikael.ronnqvist@nhh.no

Jef Van Meensel

Centre for Agricultural Economics jef.vanmeensel@ilvo.vlaanderen.be

Jaime San Martín

University of Chile

jsanmart@dim.uchile.cl

University of Chile aweintra@dii.uchile.cl

Preface

Operations Research/Management Science (OR/MS) approaches have helped people for the last 40 years or so, to understand the complex func-tioning of the systems based upon natural resources, as well as to manage this type of systems in an efficient way The areas usually viewed within the natural resources field are: agriculture, fisheries, forestry, mining and water resources

Even though, the above areas are usually viewed as separate fields of study, there are clear links and relations between them In fact, all of them share the common problem of allocating scarcity along time in an optimal manner The scale of time or length of the planning horizon is very different Thus, we have almost a continuous renewal in the case of the fisheries, periodic cycles in the case of agriculture and forestry (ranging from some few months in the case of a horticultural crop to more than a century for some forest species), and enormous periods of time much beyond the human perception in the case of mining resources But in all the cases, the key matter is to obtain an efficient use of the resource along its planning horizon Another element of connection among the different natural resources is due to the interaction between the use of the resource, and the environmental impact caused by its extraction or harvest This type of interaction implies additional complexities in the underlying decision-making process, making the use of OR/MS tools especially relevant

The above views are corroborated by the massive use of quantitative approaches in the management of natural resources It can be said that this broad field was one of the first where the OR/MS discipline was successfully applied

The papers presented correspond to invitations made to the specialists

we considered the most distinguished in each area, and we are extremely satisfied with the positive response we obtained from them In defining the subject matters, we tried to cover comprehensively the most relevant topics

in each area, from the application point of view, as well as consideration

of the operations research techniques involved In particular, we wished to highlight the successes of the OR approach to deal with problems, which involves a conceptual view of problems, modelling of complex realities, and development of algorithms for problems increasingly difficult to solve Issues of large scale, uncertainty, multiple objectives appear increasingly in these decision processes Also, we view the integration in multidisciplinary approaches, where specialists in the specific areas need to interact with operations research specialist, and the need to incorporate information tech-nologies for implementations is also present

The set of papers compiled in this volume attempts to provide readers with significant OR/MS contributions in each one of the applied areas previously defined In this way, we hope to encourage the use of quantitative techniques in order to manage the use of the different natural resources efficiently from an economic as well as an environmental point of view The papers are divided by area of application: agriculture, fisheries, forestry and mining

Acknowledgments

The preparation of the volume was a very long process that exceeds considerably the initial target Hence, we thank all the authors for their co-operation and patience All the papers were assessed following a blind reviewing process Our gratitude to all the anonymous referees

The following funding is acknowledged The work of Carlos Romero was supported by the Spanish “Ministerio de Educación y Ciencia” under research grant SEJ2005-04392 The work of Rafael Epstein and Andres Weintraub was supported by Nucleo Milenio “Complex Engineering Systems”

AGRICULTURE

In the area of agriculture we have eight chapters with different concerns such

as conceptual problems related with risk analysis, the interaction between agriculture and the environment, water resources planning, agroforestry sys-tems management, simulation of effects on agriculture of changes in the common agriculture policy, and so on OR/MS techniques used are basically the following: linear programming, multi-objective fractional programming, goal programming, multi-attribute utility theory and control dynamic optimi-zation

The chapter “Importance of whole-farm risk management in agriculture”,

by Huirne, Meuwissen and Van Asseldonk, deals with the problems ated with the definition and measurement of risk at the whole-farm level The conceptual framework is tested through a questionnaire survey among livestock and arable farmers in the Netherlands

associ-The chapter “Dealing with multiple objectives in agriculture”, by Hayashi, presents state-of-the-art of multiple criteria decision-making approaches app-lied to the selection problems in agricultural systems The chapter pays special attention to matters related with attributes definition and problem struct-uring, in order to build suitable models for agri-environmental decision-making

The chapter “Modelling multifunctional agroforestry systems with mental values: Dehesa in Spain and woodland ranches in California”, by Campos, Caparrós, Cerdá, Huntsinger and Standiford, deals with modelling agroforestry systems (“dehesas”) with the help of optimal control techni-ques Two studies, one in California and the other one in Spain, are accom-plished under a comparison basis

environ-The chapter “Environmental criteria in pig diet formulation with objective fractional programming”, by Peña, Castrodeza and Lara, incorpor-ates environmental criteria in pig diet formulation The proposed model is satisfactorily solved by resorting to an interactive multigoal programming model that allows the incorporation of goals of fractional nature

multi-The chapter “Modelling the interactions between agriculture and the environment”, by Zekri and Boughanmi, reviews the integration of different OR/MS approaches for modelling the interaction between agriculture and the environment In this way, the authors propose a decision support system based upon multi-criteria techniques and geographical information systems within a participatory decision-making perspective

The chapter “MCDM farm system analysis for public management of irrigated agriculture”, by Gómez-Limón, Berbel and Arriaza, proposes a multi-criteria approach to assist policy decision-making on water manage-ment for irrigated agriculture The methodology is a hybrid between multi-attribute utility theory and goal programming The methodology is applied to several Spanish case studies within the recent European Water Framework Directive

The chapter “Water public agencies agreeing to a covenant for water transfers: How to arbitrate price-quantity clauses”, by Ballestero, deals with inter-basin water covenants guided by the principle of arbitration and imple-mented through public agencies The methodology is illustrated with the help of a realistic example in a maritime region near the Mediterranean Sea Finally, the chapter “Positive mathematical programming for agriculture and environmental policy analysis: Review and practice”, by de Frahan, Buysse, Polomé, Fernagut, Harmignie, Lauwers, van Huylenbroeck and van Meensel, introduces a farm-level sector model, called SEPALES, based upon the approach known as a positive mathematical programming After this, SEPALES is used to simulate several economic and environmental effects on Belgium agriculture, due to some possible changes in the European Common Agricultural Policy

IMPORTANCE OF WHOLE-FARM RISK

MANAGEMENT IN AGRICULTURE

Ruud Huirne, Miranda Meuwissen, and Marcel Van Asseldonk

Institute for Risk Management in Agriculture, Wageningen University, The Netherlands

received little attention in Europe Research indicates that whole-farm management approaches, that is approaches in which multiple risks and farm activities are considered simultaneously, seem more efficient than ‘single risk and commodity strategies’ This chapter first gives an overview of risk management and then it discusses the results of a questionnaire survey among livestock and arable farmers in the Netherlands The survey deals with farmers’ perceptions of risk and risk-management strategies Risk-management stra- tegies include both ‘single risk’ strategies as well as strategies for simultane- ously covering multiple risks The latter are restricted to the type of strategies currently available in the Netherlands Next, opportunities for broadening the scope of risk-management strategies covering multiple risks are dis- cussed The paper concludes by identifying areas for further research in the field of whole-farm risk management

risk-Keywords: Risk management, agriculture, whole-farm approach, multiple risks,

question-naire survey

1 INTRODUCTION

The agricultural firm is constantly developing The farm is and remains an essential player in the agricultural supply chain and in the rural area The differences between the agricultural sector and the rest of the industry are getting smaller and smaller Increasing farm sizes result in a more indus-trialized way of undertaking such operations Important ‘new’ characteristics

of such bigger, industrialized farms include: importance of manufacturing processes (vs commodities); a systems approach to production and distri-bution; separation and realignment of the stages in the food chain for the purpose of efficiency and low cost-price; negotiated coordination among

those stages and with the environment (rural area); concern about system power and control; and new kinds of risk combined with a more important role for information This implies that risk considerations are becoming more important and should be addressed in a more formal way

Income from farming is usually considered rather volatile because of a whole series of stochastic factors, that is risk Over the years, a range of risk-management strategies has been used to reduce, or to assist farmers to absorb, some of these risks (see later) Risk-management strategies, especially risk-sharing strategies, generally deal with only one type of risk at a time For instance, futures market contracts deal with price risks, hail and storm insurance schemes cover weather-related production risks, and livestock insu-rance schemes cover the death of animals Even disaster relief programs in such events as droughts and floods consider only one type of risk (which, in itself, is relevant if the whole – or a notable part of the – crop or herd is destroyed)

This chapter first discusses risk management in general (definition, sources

of risk, risk-management strategies) and then the results of a questionnaire survey among livestock and arable farmers in the Netherlands Because Dutch farms are not really representative compared to farms in many other countries, the results of the survey should be seen as an example The survey deals with farmers’ perceptions of risk and risk-management strategies Risk-management strategies include both ‘single risk’ strategies as well as strategies for simultaneously covering multiple risks The latter are restricted to the type of strategies currently available in the Netherlands Next, opportunities for broadening the scope of risk management strategies covering multiple risks are discussed The chapter concludes by identifying areas for further research in the field of whole-farm risk management

The concepts of ‘risk’ and ‘uncertainty’ have already been referred to several times It is time to elaborate upon them The meanings of ‘risk’ and

‘uncertainty’ come close (Hardaker et al., 2004) Uncertainty is the result of

incomplete knowledge Risk can be defined as uncertain consequences or results at the time of making decisions Risk particularly concerns exposure

to unwanted, negative consequences Risk management concerns the way

in which managers deal with risk and uncertainty (Meuwissen et al., 1999, 2001; Huirne et al., 2000; Van Asseldonk et al., 2001; Huirne, 2002)

2.1 Types of Risk

The current government policy has increasingly been aimed at creating an open market system This results in, amongst other things, the fact that agriculture in the Netherlands is increasingly confronted with price-making

in international markets, such as the world market, which generally means

lower and definitely more fluctuating prices (Huirne et al., 1997; Meuwissen

et al., 1999) Further modernization of the sector has resulted in increasing

economic consequences Dealing with such risks, that is risk management, is gaining more and more importance, not only for individual farmers, but for all firms in the agricultural supply chain

Many activities of an agricultural firm take place outdoors and are dependent The agricultural sector also deals with live material Because of this the sector is an outstanding example being exposure to risks (Anderson

weather-et al., 1977; Barry weather-et al., 2000; Van Asseldonk weather-et al., 2001; Hardaker weather-et al.,

2004) Production risks are caused by the unpredictable character of the weather and hence uncertainty as to the physical yield of animals and crops Diseases and infestations can have a great influence on farm results, as the classical swine fever outbreaks in 1997/1998 and the foot-and-mouth disease outbreaks in 2001 clearly showed

Moreover, the prices of production means most often purchased (such

as concentrates, fertilizer, pesticides and machines) and of products sold (such as milk, tomatoes and cut flowers) are not known, at least not at the time decisions on these have to be taken As already mentioned, farmers are increasingly exposed to price-making forces in unpredictable markets Thus, market and price risks are important factors

Governments form another source of risk to farmers Changes in laws and regulations with respect to running the farm can have far-reaching consequences for farm results Examples are the continuing changes in the regulations regarding environment, pesticides, animal diseases and animal welfare On the other hand, governments have also set off particular risks (up to now)

Farmers working on their farms are themselves a risk to the profitability and continuity of the farm The farm’s survival may be threatened by death

of the owner, or by divorce of a couple together running the farm term illness of the owner or employees can also cause considerable losses

Long-or can increase the costs considerably Such risks are called human Long-or sonal risks

per-There are also financial risks involved (Belli et al., 2001) These are

related to the financing of the farm Using borrowed capital (such as mortgages and the like) means that first the interest needs to be paid before increasing one’s equity capital For farms with relatively much debt capital (for example, as a result of large investments), little will be left as a reward

to one’s equity capital at times of high interest rates Only farms that are entirely equity-financed are not subject to such financial risks, but yet can sustain capital loss Other risks connected to the use of credit and loans are uncertain interest rates and inability to obtain a loan or mortgage

2.2 Reducing and Sharing Risk

Risks are thus unavoidable and influence almost any decision the farmer takes That is to say risks are present, but can be counteracted The farmer should anticipate such risks by his management But how? In what way can risks be reduced? There are two categories of measures to reduce risks:

taking measures within the farm and sharing risks with others (Huirne et al., 1997; Belli et al., 2001; Huirne, 2002; Hardaker et al., 2004)

During many uncertain events (extra) information can be obtained easily For example, asking for the weather forecast, analyzing feed or soil samples and consulting experts Also particular risks can possibly be avoided or prevented It is known that certain activities carry more risks than others Reducing farm contacts can, for example, reduce the risk of disease intro-duction considerably Another good strategy to minimize risks is not to invest all of one’s money on a single farm activity By selecting a combi-nation of activities, risks can be considerably reduced The same holds for having various suppliers and buyers Flexibility can be mentioned as a last measure at the farm level Flexibility refers to how well a farm can anticipate changing conditions For example, by investing in multipurpose machines and buildings

The second set of measures refers to sharing risks with others (Huirne

et al., 1997; Hardaker et al., 2004) One possibility here is buying insurance

At present, there are several types of insurance available, with which, by payment of a premium, risks can be reduced or even eliminated The farmer can also conclude contracts for example with suppliers and buyers in which price agreements are laid down Agreements can be made on the duty to deliver and to buy as well as on the quality of the products or raw materials Lastly, by using the futures market, price risks can largely be eliminated The futures market is not yet well known in the Netherlands, but in the USA

it is popular for a number of agricultural products

Most farmers try to reduce risks when they face decisions that may have

a considerable influence on their income or wealth (Anderson et al., 1977; Belli et al., 2001; Hardaker et al., 2004) Examples of such decisions are

sizeable investments in milk quotas or in a second farm enterprise The attitude of reducing exposure to risks is called risk aversion A risk-averse person is willing to sacrifice part of his income to reduce risks This consi-deration serves as a means to make a choice among the above measures However, reducing risks will generally involve a cost

2.3 Risk Perception

Managers, policy makers and researchers alike often have a binary way

of dealing with risk and uncertainty One either assumes certainty and an exactly predictable future, or uncertainty and an entirely unpredictable future In the latter case further analyses are often omitted and decisions are made either intuitively or not made at all Under- as well as overestimating the risks is potentially dangerous Further analysis reveals that there are at

least four levels of risk and uncertainty (Courtney et al., 1997):

1 A clear-enough future; a single forecast precise enough for the purpose of decision making

2 Alternate futures; a few discrete outcomes that define the future

3 A range of futures; a whole range of possible outcomes

4 True ambiguity; no basis to forecast the future

Levels 1 and 4 do not occur very often in practice; they are extreme situations Therefore, it is all the more distressing that many managers and advisors regularly operate at these levels of risk Particularly working at level 1 where calculations are carried out and advice is given under the assumption of complete information and certainty, is alarming

MANAGEMENT

3.1 Materials

The questionnaire survey included questions on: (i) the farm, (ii) farmers’ risk attitude, (iii) farmers’ perception of risk-management strategies, (iv) their perceptions of risks and the extent to which risks are managed on their own farm, (v) farmers’ ability to define ‘risk management’, and (vi) farmers’ interest into assistance for setting up a whole-farm risk-management plan

for their own farm Most questions were closed questions, mainly in the form of Likert-type scales ranging from 1 to 5 (Churchill, 1995) In total, the questionnaire included 177 variables The (pretested) questionnaire was sent in July 2001 to 390 clients of the Rabobank (major agricultural bank

in the Netherlands) These included cattle, pig, poultry and arable farmers After screening on completeness, the questionnaires of 101 farmers were available for statistical analyses, that is, the effective response rate was 26%

I am willing to take more risks

than other farmers 7 16 44 22 11 3.14 1.04

I need to take risks to be

I am reluctant to introducing

New technologies first need to

I am more concerned about

losses than forgoing some

From the scores in Table 1 it can be concluded that based on these tions respondents have a risk-seeking attitude It is noteworthy that this holds for all statements

ques-Table 2 shows farmers’ perceptions of risk-management strategies We subdivided the strategies into strategies that cover single risks and strategies that simultaneously cover multiple risks In making this subdivision we assumed that new technologies are primarily implemented to deal with production risks, that leasing machinery has mainly to do with financial risks and that leasing milk quota mostly deals with production risks In the category ‘multiple risk strategies’, we assumed that vertical and horizontal cooperation deal with both price and production risks In relation to spatial diversification we supposed that this has not only to do with diversifying production risks but most likely also with diversifying institutional risks (e.g

in case of environmental requirements) and/or price risks

Table 2 shows that, in general, farmers perceive the single ment strategies as more relevant than the strategies covering multiple risks:

risk-manage-of the ten strategies ranked highest (see last column ‘overall rank’) only four strategies are within the multiple risk category These strategies include increasing the solvency rate, comprising financial reservations, on-farm diver-sification and vertical cooperation Popular risk-management strategies in

‘single-risk strategies’ are strict hygiene rules, business insurance, personal insurance and the application of new technologies

Table 2 Perception of risk-management strategies, n = 101 (1: not relevant at all; 5: very

Application of new technologies 3.64 0.93 6

Price contracts for farm input 2.90 1.10 12

Price contracts for farm output 2.88 1.10 13

Comprise financial reservations 3.81 0.99 3

Asking respondents for their ‘top 3’ risk-management strategies resulted

in the following answers (the percentage of respondents indicating a cular strategy is given in parentheses):

parti-1 Increase solvency rate (16%), on-farm diversification (16%), comprise financial reservations (14%);

2 Increase solvency rate (11%), comprise financial reservations (10%), strict hygiene rules (10%); and

3 Vertical cooperation (14%), on-farm diversification (12%), application of new technologies (12%)

From these answers it can be seen that from the ‘top 4 strategies’ from Table 2 (i.e 1: strict hygiene rules; 2: increase solvency rate; 3: comprise financial reservations; 4: business insurance) multiple risk strategies (option

2 and 3) are favourite in among top 3

Table 3 illustrates farmers’ perceptions of risks and the extent to which they believe that the risks are dealt with on their own farm There are seven

risk categories Besides the ones distinguished by Hardaker et al (2004) we

added the categories ‘liability risks’ and ‘risks related to immovable objects’

Table 3 Perception of risk and the extent to which risks are managed on own farm, n = 101

Relevance of risk (1: not relevant; 5: very relevant)

Risk is managed on my farm

(%) Average Std Overall

rank

yet

Yes partly

Dutch suppliers or

buyers

3.50 1.21 11 42 13 33 12 –

1Not applicable

provisions because of

declining farm values

2.61 1.13 21 28 14 16 42 –

Table 3 shows that farmers perceive production and price risks as very important Liability risks and financial risks are ranked 6th and 7th respec-tively With respect to the management of the risks, farmers are convinced that they (largely) handled the production risks, institutional risks (as far

as it concerns governmental regulations), personal risks, risks related to immovable objects, liability risks and financial risks Note that for some type of risks the numbers in the column ‘yes I managed the risk partly’ are higher than for other risks This is for instance the case for liability risks Not (yet) adequately dealt with are price risks, risks related to the elimi-nation of government support (e.g in case of droughts and livestock epidemics) and the decrease in farms’ collateral value

The two remaining parts of the questionnaire, that is farmers’ ability to define risk management and farmers’ interest into assistance for setting up a whole-farm risk management-plan for their own farm, led to the following results:

1 About 70% of the respondents was able to adequately define risk management

2 About 62% of the respondents showed interest in assistance for loping a risk management plan for their own farm

deve-There was a significant positive relationship between farmers being able

to define risk-management and those interested in a risk management plan (P ≤ 0.05)

The ‘multiple risk strategies’ included in the chapter so far are classical examples of on-farm diversification, off-farm employment, increasing the solvency rate, etc Vertical and horizontal cooperation are more recent examples (Boehlje and Lins, 1998) This section discusses three further oppor-tunities for simultaneously covering multiple risks: certification, revenue insurance and stabilization accounts Certification can be categorized as an

‘on-farm strategy’; revenue insurance and stabilization accounts are sharing strategies’

‘risk-Certification is already widely available in the Netherlands Examples include KKM (Chain Quality Milk) for dairy farms, PVE/IKB (Integrated Chain Control) for pig farms, Safe Quality Food for primary producers (SQF-1000) and Good Agricultural Practices as defined in Eurep-GAP

Certification reduces production risks (through, among others, improved internal efficiency and less failure costs), liability risks (since certification effectuates due diligence) and price risks – if markets for certified products

are more stable than other markets (Unnevehr et al., 1999; Velthuis et al., 2003; Meuwissen et al., 2003b)

Revenue insurance is not (yet) available in the Netherlands It taneously covers price and yield risks of a particular commodity If the correlation between both parameters is negative (i.e lower yields result in higher prices, and vice versa) revenue insurance should be less expensive than insurance for yields only The concept has existed in the USA for many years (see for instance Goodwin and Ker, 1998) Schemes are highly subsidized by the US government (Skees, 1999) However, since these type

simul-of insurance schemes seem legitimate in the WTO-framework (i.e they fit into the “green box” representing allowed forms of support), the European

Commission is now considering similar tools (Meuwissen et al., 2003a)

Stabilization accounts not only cover multiple risks but (if relevant for a particular farm) also multiple commodities The principle of stabilization accounts is that farmers can put money into the account in high-income years (when marginal taxes are high) while withdrawing it in low-income years (when marginal taxes are low) Examples of stabilization accounts (curr-ently not available in the Netherlands) include the Canadian Net Income Stabilization Accounts (NISA) and the Australian Farm Management Deposits NISA is a whole-farm program in which farmers put money into a bank account, government matches the farmer’s deposits (“dollar for dollar”), and each farmer can withdraw from the account in adverse times Also NISA

is legitimate under WTO regulations The Canadian government is currently reconsidering their program in order to also include on-farm food safety issues and environmental programs The Australian scheme equals the Canadian one but without the matching contributions from the government (European Commission, 2001)

This chapter was set up around ‘whole-farm risk management’ Results from the questionnaire indicate that farmers perceive that they have managed their farm risks quit well (with some exceptions, mainly in the field of price risks and risks related to the elimination of government support) Farmers generally prefer ‘single-risk and commodity strategies’

Following a whole-farm risk-management approach, that is an approach

in which multiple risks and farm activities are considered simultaneously, may be more efficient, but probably also more complicated Of the respon-dents 62% indicated that they were interested in assistance in setting up a whole-farm risk-management plan This percentage may even have been higher if respondents had not known that the survey was initiated by the Rabobank (which has some direct interest in such risk-management plans) The multiple risk strategies discussed (i.e certification, revenue insu-rance and stabilization accounts) have some features of a whole-farm risk-management approach For instance, when designing revenue insurance schemes it is relevant to have insight into the correlation between prices and yields When setting up (subsidized) stabilization accounts, insight is needed into the correlation of revenues among various farm activities Certification programs require the identification of critical control points of a farm, for example with respect to food safety

From the above, we define the following areas for further research in the field of whole-farm risk management:

1 An analysis of (the dynamics in) correlations between prices and yield of various agricultural activities

2 An analysis of the critical control points of a farm from the perspective of the overall farm viability

After these steps have been taken, whole-farm risk-management plans can be designed – and the ideal partners for advising about them can be identified

6 REFERENCES

Anderson, J.R., Dillon, J.L., Hardaker, J.B., 1977, Agricultural Decision Analysis Iowa State

University Press, Ames, Iowa

Barry, P.J., Ellinger, P.N., Hopkin, J.A., Baker, C.B., 2000, Financial Management in Agriculture Interstate, Danville, Illinois

Belli, P., Anderson, J.R., Barnum, H.N., Dixon, J.A., Tan, J.-P., 2001, Economic Analysis of Investment Operations WBI Development Studies, The World Bank, Washington, D.C

Boehlje, M.D., Lins, D.A., 1998, Risks and risk management in an industrialized agriculture

Agric Finance Rev 58: 1–16

Churchill, G.A., 1995, Marketing Research Methodological Foundations The Dryden Press,

New York

Courtney, H., Kirkland, J., Viguerie, P., 1997, Strategy under uncertainty Harvard Bus Rev

75: 67–79

European Commission, 2001, Risk Management Tools for EU Agriculture, with a Special Focus on Insurance Working Document, Agriculture Directorate-General

Goodwin, B.K., Ker, A.P., 1998, Revenue insurance—a new dimension in risk management

Choices 13(4): 24–27

Hardaker, J.B., Huirne, R.B.M., Anderson, J.R., Lien, G., 2004, Coping with Risk in Agriculture, second edition CAB International, Wallingford

Huirne, R.B.M., 2002, Strategy and risk in farming Neth J Agr Sci 50: 249–259

Huirne, R.B.M., Hardaker, J.B., Dijkhuizen, A.A (eds), 1997, Risk Management Strategies

in Agriculture: State of the Art and Future Perspectives Mansholt Studies, No 7,

Wageningen Agricultural University, Wageningen

Huirne, R.B.M., Meuwissen, M.P.M., Hardaker, J.B., Anderson, J.R., 2000, Risk and risk

management in agriculture: an overview and empirical results Int J Risk Assess

Manage 1: 125–136

Meuwissen, M.P.M., Huirne, R.B.M., Hardaker, J.B., 1999, Income Insurance in European Agriculture Scientific Report EU-Project, European Economy, No 2, DGII, Brussels,

pp 95

Meuwissen, M.P.M., Huirne, R.B.M., Hardaker, J.B., 2001, Risks and risk management

strategies; an analysis of Dutch livestock farmers Livest Prod Sci 69: 43–53

Meuwissen, M.P.M., Huirne, R.B.M., Skees, J.R., 2003a, Income insurance in European

agriculture EuroChoices (Spring), 12–17

Meuwissen, M.P.M., Velthuis, A.G.J., Hogeveen, H., Huirne, R.B.M., 2003b, Traceability

and certification in meat supply chains J Agribusiness 21: 167–181

Skees, J.R., 1999, Agricultural risk management or income enhancement? Regulation Cato

Rev Bus Gov 22: 35–43

Unnevehr, L.J., Miller, G.Y., Gómez, M.I., 1999, Ensuring food safety and quality in

farm-level production: emerging lessons from the pork industry Am J Agric Econ 81: 1096–

1101

Van Asseldonk, M.A.P.M., Meuwissen, M.P.M., Huirne, R.B.M., 2001, Stochastic simulation

of catastrophic hail and windstorm indemnities in the Dutch greenhouse sector Risk Anal

21: 761–769

Velthuis, A.G.J., Unnevehr, L.J., Hogeveen, H., Huirne, R.B.M (eds), 2003, New Approaches

to Food-Safety Economics Kluwer Academic Publishers, Dordrecht, The Netherlands

DEALING WITH MULTIPLE OBJECTIVES

IN AGRICULTURE

Kiyotada Hayashi

National Agricultural Research Center, 3-1-1 Kannondai, Tsukuba, Ibaraki 305–8666, Japan

problems in agricultural systems It also discusses life cycle impact assessment (LCIA) from the decision analytic framework Special attention is paid to the attributes (evaluation criteria) used for evaluating agricultural systems by considering their impacts on the environment Problem structuring in decision analysis, which is related to the definition of impact categories in LCIA, is also discussed to construct multiple objectives suitable for agri-environmental decision making

Keywords: Agricultural systems, sustainability, multicriteria analysis, selection problems,

life cycle impact assessment (LCIA), category indicators

1 INTRODUCTION

The interactions between agriculture and the environment are causing major public concern over the appropriateness of modern agricultural practices (Pretty, 2002) For example, intensive agricultural practices that make excessive use of chemical fertilizers and manure have negative impacts on the environment and are therefore recognized as the cause of agricultural nonpoint pollution Moreover, modern intensive agriculture is threatening wild biodiversity Since agriculture is now spreading to the remotest parts of the world in destructive forms, wild biodiversity has been reducing, thus undermining the sustainability of the food production system (McNeely and Scherr, 2003)

As a result, operations research for natural resource management has become important in assisting farmers and extension specialists decide whether to introduce alternative practices to make agriculture more sustainable

and in helping policy makers judge the appropriateness of policies designed

to remedy agri-environmental issues Multicriteria analysis can be a typical

methodology used for these purposes

However, in most case studies, attention is centered on the relative

per-formance of several agricultural systems and the selection of an appropriate

agricultural system This means that the problems pertaining to how

evalua-tion criteria (attributes used for multicriteria analysis) should be constructed

have not been sufficiently discussed, although the consideration of these

pro-blems is indispensable in the effective performance of multicriteria analysis

This is because the result of the analysis may be inadequate to ensure

advan-tageous behavior unless the problem is structured appropriately

Therefore, this chapter reviews the multicriteria analysis applied to analyze

the relative performance of agricultural systems Special attention is paid to

the environmental impacts of agricultural practices Life cycle impact

assess-ment (LCIA) is also included in the following survey and is discussed from

the perspective of multicriteria analysis Considering the appropriateness of

multicriteria analysis and related methods for resolving agri-environmental

problems will offer food for thought on problem structuring in decision

ana-lysis for agriculture, through clarifying the issues on how to set appropriate

multiple objectives in agricultural decision making

Section 2 reviews the multicriteria analysis applied to the problems in

selecting agricultural systems, in which the environmental impacts of

agri-cultural systems are evaluated, and discusses the problems with weighting

In Section 3, the applications of LCIA to agricultural production are

sur-veyed from the perspective of multicriteria analysis, after presenting a

tripartite classification of methodology, which consists of the direct

appli-cation of multicriteria analysis, multicriteria analysis using the midpoint

approach to LCIA, and multicriteria analysis using the end point approach

to LCIA In the subsequent section, some related topics are presented to

better understand the applicability of multicriteria analysis to natural resource

management

OF AGRICULTURAL SYSTEMS

This section describes the selection problems in agricultural systems, and

agricultural systems are defined as discrete alternatives and are selected (or

ranked) with respect to multiple attributes Although multicriteria analysis

contains multiobjective planning (multiobjective mathematical programming, including goal programming), this chapter focuses on the methods for selecting discrete alternatives (See Hayashi (2000) and Romero and Rehman (2003) for information on multicriteria analysis, including multiobjective mathematical

programming, and see Hardaker et al (2004) for information on decision

ana-lysis in agriculture.)

Table 1 illustrates the models used for evaluating agricultural systems by considering their impacts on the environment In these applications, both compensatory and noncompensatory methods are used The former methods use multiattribute value functions or compromise programming, and the latter methods apply the concept of outranking (see, e.g., DETR, 2000) Since it is difficult to elicit attribute weights from decision makers (the farmers or experts), only ranking information is used in the applications

of the former methods For example, to obtain the best and worst overall

values, Yakowitz et al (1993) define the following mathematical program:

1minimize or maximize ( ) ( )

n

i j j ij j

w

w ≥w ≥w ≥w ≥

where v(a i ) is the overall value of alternative a i , w j is the weight assigned to

the jth attribute, v j (·) is the value function for the jth attribute, x ij is the jth

attribute level for alternative a i , and n is the number of attributes

Table 2 lists the attributes (criteria) used for analyzing the selection problems in agricultural systems Since these are examples of evaluating agricultural systems on the basis of environmental impacts, attention is paid

to the trade-offs between economic objectives and environmental objectives, with the exception of Arondel and Girardin (2000) This implies that the problems can be depicted graphically as value trees (objectives hierarchies) that contain profitability and the environmental quality of soil and water Conducting multicriteria analysis for the selection problems provides us with an integrated perspective on the interaction between agriculture and the environment; thus, multicriteria analysis supports farmers’ and policy makers’

Table 1 Models used for selecting an agricultural system

Yakowitz et al (1993) MAVFb Ranking c A farmer d

Foltz et al (1995) MAVF b Pairwise ranking e A farmer

Tiwari et al (1999) Compromise

pro-(2000) Outranking (ELECTRE TRI) The revised “weighting with cards” method Experts

Strassert and Prato

(2002)

Outranking (balancing and ranking)

a References are restricted to referred journals in English

b The multiattribute value function

c The same method is applied

d The explicit explanation of the decision maker is not provided

e Kirkwood and Sarin (1985)

decisions on whether to introduce alternative agricultural systems However,

there are two difficulties in applying this methodology One difficulty is the

problem of weighting Since the meaning of weights is dependent on models,

weight parameters may have widely differing interpretations for different

methodologies and different decision contexts (Belton and Stewart, 2002) In

multiattribute value (utility) functions, which have clear algebraic meanings

of attribute weights as compared with other methodologies such as

outrank-ing methods, weight elicitation methods that do not rely on attribute ranges

might lead to biased weights (Von Nitzsch and Weber, 1993; Fischer, 1995;

Belton and Stewart, 2002) Nevertheless, several of the applications listed

in Table 2 elicit attribute weights without referring to attribute ranges This

difficulty with weighting is a common and serious mistake in the application

of multicriteria analysis in various research fields, and this has already

been realized in some application areas; for example, in the integration of

geographical data by Geographical Information Systems (GIS), the range

problem just mentioned has been recognized as a common source of error

(Malczewski, 1999)

The other difficulty is related to problem structuring, the importance of

which has been stressed recently Since, for example, the trade-offs between

nitrate concentrations in surface water and atrazine concentrations in

per-colation are difficult to understand for decision makers and even for experts,

Table 2 Attributes used for selecting an agricultural system

Soil loss d

Heilman

et al

(1997)

Net returns N (runoff )

NO 3 –N (percolation) Atrazine (runoffAtrazine )

(sediment) All other pesticides in surface or groundwater

Soil detachment Sediment yield

Pesticide management (amount, half- life, mobility, toxicity, location, date)

Water management (hydric balance, amount)

Atrazine (application) [as drinking water quality]

Soil erosion

a See Table 1

b Estimated by the Erosion Productivity Impact Calculator (EPIC)

c Estimated by Groundwater Loading Effects of Agricultural Management Systems (GLEAMS)

d Calculated by the Universal Soil Loss Equation (USLE)

e The USDA Natural Resources Conservation Service

it is necessary to introduce a methodology for transforming the evaluation

data into other values to make the meaning easily comprehensible This

procedure can be considered as impact assessment Although LCIA was

developed in a research field different from decision analysis, it can be

recognized as a type of multicriteria analysis Hence, the subsequent section

reviews the LCIA applied to agricultural production

OF MULTICRITERIA ANALYSIS

As a preparation for discussing the earlier applications of LCIA in

agri-culture, this section first describes the trichotomy that consists of (1) the

direct application of multicriteria analysis, (2) multicriteria analysis using

midpoint approaches to LCIA, and (3) multicriteria analysis using end point

approaches to LCIA In the first category (Fig 1), the inventory data are

directly transformed into environmental indicators (overall values) by

weight-ing Although the system boundary of LCIA in general includes fertilizer and

pesticide production processes, these figures depict only the direct impacts

of agricultural practices In addition, although the figure lacks the economic

criteria that the applications in the previous section have, it is possible to add

the criteria that measure economic performance

Figure 1 The direct application of multicriteria analysis A rectangular node depicts an

agricultural practice, which can be considered as a decision because the decision maker can

control it directly A rounded rectangular node means any intermediate concept or variable A

hexagonal node depicts an overall value or indicator, which can be used to evaluate the

relative desirability of agricultural practices or agricultural systems (Adapted from Hayashi

and Kawashima, 2004.)

Figure 2 Multicriteria analysis using a midpoint approach to LCIA The classification is

based on Guineé (2002) See Fig 1 for the notation Adapted from Hayashi and Kawashima (2004)

The second category (Fig 2) introduces the category indicators at the midpoint level, such as global warming potential (GWP), which correspond

to the attributes in multicriteria analysis In other words, inventory data are transformed into impact categories using, for example, GWP Although the term “normalization” is used in the figure as is often the case with LCIA, it

is recognized as the construction of value functions from the multiattribute value function framework

In the third category (Fig 3), inventory data are integrated into more comprehensive concepts such as human health and ecosystem quality, which are the category indicators at the end point level Fate analysis, exposure and effect analysis, and damage analysis are used to derive the category indi-cators at this level (For a discussion on the relationship between decision analysis and life cycle assessment (LCA), see Miettinen and Hämäläinen,

1997; Hertwich and Hammitt, 2001a, b; and Seppälä et al., 2002)

Table 3 illustrates the applications of LCA to the assessment of the ronmental impacts of agricultural systems Examples include arable farming, milk production, and animal production Most of the applications define the decision alternatives, although the problems in those applications are not explicitly formulated as decision problems As for the functional unit, there are the unit weights of products and the unit area of production

envi-Table 3 Examples of LCA in agriculture (definition of the problem)

Units

Hanegraaf

et al (1998)

Energy crop production in the Netherlands

Route+Crop (GAP) 1 GJ and 1 ha

Haas et al

(2001)

Grassland ming in Germany

far-Intensive, extensive, and organic farming

1 ha and 1 t milk Brentrup

et al (2001a, b)

Sugar beet production in Germany

Sugar beet production with calcium ammonium nitrate (solid fertilizer), urea (solid fertilizer), and urea ammonium nitrate solution (liquid fertilizer)

1 t of extractable sugar

Eide (2002) Industrial milk

production in Norway

Small, middle-sized, and large dairy

1000 L of drinking milk brought to the consumers

Bennet et al

(2004)

GM sugar beet production in the UK and Germany

Conventional and tolerant sugar beet

GM-herbicide-50,000 kg fresh weight

of sugar beet

Brentrup

et al (2004)

Winter wheat production in the UK

(Nitrogen fertilizer rate) 1 t of grain

Soil cultivation, open, and closed hydroponic systems (+3 waste management scenarios)

1 kg of tomatoes

Source: Hayashi et al 2006

a The references are restricted to referred journals in English Papers that analyze only food processing

are not included Web of Science was used to search the papers

Figure 3 Multicriteria analysis using an end point approach to LCIA Cause and effect

relationships in fate analysis, exposure and effect analysis, and damage analysis are based on

Jolliet et al (2003), although the relationships regarding global warming are based on Itsubo

and Inaba (2003) (Adapted from Hayashi and Kawashima, 2004.)

agriculture The commonly used categories are climate change, human

toxi-city, acidification, and eutrophication These categories can be recognized

as attributes in multiattribute value functions, although weighting is not

Note: The column headings are as follows: (1) resource use, (2) land use, (3) climate change,

(4) human toxicity, (5) ecotoxicity, (6) acidification, (7) eutrophication, (8) stratospheric ozone

depletion, (9) photo-oxidant formation and (10) others

a The references are restricted to referred journals in English Papers that analyze only food processing

are not included

b Minerals to soil and water, pesticides, soil erosion, groundwater depletion, waste production and

utilization, contribution to biodiversity, contribution to landscape values

c Soil functions, water quality, and biodiversity

d Pesticide use

e Water resource

performed except in the work of Brentrup et al (2001) in which

Eco-Table 4 summarizes the impact categories used in the applications to