Exploring Environmental Change Using an Integrative Method

Exploring Environmental Change Using an Integrative Method Environmental Problems and Social Dynamics A series of boo...

Author: Mark Lemon

113 downloads 1240 Views 10MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

Exploring Environmental Change Using an Integrative Method

Environmental Problems and Social Dynamics A series of books edited by Peter M.Allen, Cranfield University, Cranfield, UK and Sander E.Van der Leeuw, Université de Paris, Paris, France Volume 1 Cities and Regions as Self-Organizing Systems: Models of Complexity Peter M.Allen Volume 2 Environmental Management in European Companies: Success Stories and Evaluation edited by Jobst Conrad Volume 3 Exploring Environmental Change Using an Integrative Method edited by Mark Lemon

This book is part of a series. The publisher will accept continuation orders which may be cancelled at any time and which provide for automatic billing and shipping of each title in the series upon publication. Please write for details.

Exploring Environmental Change Using an Integrative Method Edited by

Mark Lemon Cranfield University, Cranfield, UK

Gordon and Breach Science Publishers Australia • Canada • China • France • Germany • India Japan • Luxembourg • Malaysia • The Netherlands Russia • Singapore • Switzerland

This edition published in the Taylor & Francis e-Library, 2005. “To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to www.eBookstore.tandf.co.uk.” Copyright © 1999 OPA (Overseas Publishers Association) N.V. Published by license under the Gordon and Breach Science Publishers imprint. All rights reserved. No part of this book may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying and recording, or by any information storage or retrieval system, without permission in writing from the publisher. Printed in Singapore. Amsteldijk 166 1st Floor 1079 LH Amsterdam The Netherlands British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library. ISBN 0-203-30403-9 Master e-book ISBN

ISBN 0-203-34358-1 (Adobe eReader Format) ISBN 90-5699-193-0 (Print Edition) ISSN 1027-2607

CONTENTS

PREFACE

vii

LIST OF CONTRIBUTORS

ix

1

Policy Relevant Research: The Nature of the Problem Mark Lemon and Roger Seaton

1

2

Towards an Integrative Method Mark Lemon and Roger Seaton

17

3

Background to Agriculture and Degradation in the Argolid Valley Mark Lemon and Nenia Blatsou

27

4

Social Enquiry and Natural Phenomena Mark Lemon

39

5

Complexity, Systems and Models Paul Jeffrey, Roger Seaton and Mark Lemon

69

6

Agricultural Production and Change Mark Lemon and Nenia Blatsou

81

7

Technology and Agricultural Production in the Argolid Mark Lemon and Nenia Blatsou

115

8

Structural Weaknesses and the Argolid? Mark Lemon and Nenia Blatsou

135

9

Perceived Uncertainty and Farming: Establishing a Framework for Crop Choice Mark Lemon and Nenia Blatsou

157

10

Policy Relevant Modelling in the Argolid: From Sociological Investigation to Crop Choice Model Roger Seaton and Ian Black

167

11

Towards a Strategic Complex Systems Model of the Water/Salt System Roger Seaton

185

12

The Strategic Model of Water Flow Peter Allen

199

13

Development of an Enhanced Integrated Dynamic Model Tim Oxley

211

vi

14

Where to From Here? Mark Lemon and Tim Oxley

227

REFERENCES

237

INDEX

245

PREFACE

The International Ecotechnology Research Centre (IERC) is a multi-disciplinary unit consisting of physicists, ecologists, engineers, educationalists, economists, archaeologists, sociologists and hybrids thereof. Over the past eight years the Centre has undertaken work ranging from the eutrophication in the North Sea through the disposal of waste by land-fill to studies of knowledge and technology transfer and organizational change. This apparently disparate history does, however, have a consistent and central theme which lies in the importance of non-linear dynamics and complex systems thinking for understanding and managing change and in particular the role of knowledge transfer in the process. While this theoretical starting point may be articulated differently by representatives of the disciplines, within the group a general scepticism about planning for ‘end states’ is matched by the recognition that no single disciplinary perspective can provide an adequate insight into naturalhuman interactions. An integrative method has been developed which is consistent with this interpretation of systems as complex but which is sufficiently flexible to recognise the need to select the relevant skills for particular pieces of research. The emergence of this method, and the selectivity within it, has been accompanied by the need for disciplinary humility, a requirement that is not always foremost in the academic tool box. Making sure that the nuts and bolts are roughly in the correct place is fundamental to representing the machine. By the same token the absence of the odd nut or bolt will not necessarily be too detrimental to that representation. Herein lies an important point. What is being sought by the group is a representation of the real world that is capable of generating questions about possible futures rather than providing solutions. Those questions need to be relevant and as such must be capable of supporting choices relating to the future—in other words they must be policy relevant. This study is an attempt to convey this integrative method through a specific context (natural resource degradation in the Argolid Valley in Greece1), however, it is intended that the process described is of a more general use and as such should not be restricted to a readership whose interests lie within agronomy or hydrology. Throughout the text background information will be provided concerning the techniques adopted and as such it is anticipated that an appreciation of the role to be played by other disciplines will develop. There are sections of the text however that are more technically demanding; this is particularly the case with the modelling chapters towards the end of the book. While the equations are not for general consumption it is hoped that these sections can be read at a more descriptive level which reinforces the underlying message of integrative method.

viii

While it is not our intention to advocate the nurturing of multi-disciplinarians, a movement towards the acquisition of a set of transdisciplinary skills is desired. It is this which can support a more informed choice about when different disciplines and techniques are useful. We would also like to acknowledge the support provided for this work by the Agricultural University of Athens, Department of Hydrology and in particular P.Giannoupoulos (Takis). The project is unlikely to have reached completion without the advice, support and at times forceful cajoling of Sander Van der Leeuw.

1This

work was funded by DGX11 of the Commission of the European Union as part of the environment Desertification Programme, Archaeomedes (EV5V-0021) and Environmental Perception (EV5V-0486) projects.

LIST OF CONTRIBUTORS

Peter Allen

is Head of Ecotechnology Research at the International Ecotechnology Research Centre (IERC) at Cranfield University, UK. He has a background in theoretical physics and was a Senior Research Fellow at the Universiti Libre de Bruxelles where he worked on self-organising systems with the Nobel Prize winner, Ilya Prigogine. His research has focused on the mathematical modelling of change and innovation in social, economic and ecological systems, and the development of integrated systems models which link the physical, ecological and socio-economic aspects of complex systems as a basis for improved decision support systems, Ian Black is a Senior Lecturer in the Cranfield University School of Management. He is an economist with an interest in the use of mathematical models to explain producer and consumer behaviour. He is also involved in research concerning various policy initiatives in the European Union and how to design effective instruments to achieve sustainable development. Nenia Blatsou is a graduate of the Agricultural University of Athens, Department of Horticulture and has worked for the Greek public sector as an agronomist. She undertook most of the field work for the Argolid project and is currently employed as a researcher in the IERC with specific interests relating to sustainable agricultural land use in the Mediterranean and farmer’s decision-making. Paul Jeffrey is a Research Officer who recently rejoined the IERC following two years as an International Post-Doctoral Fellow based at the Hebrew University in Jerusalem. He has an interdisciplinary background with research interests in the relationship between technology and society, the co-evolutionary dynamics of resilient communities and the study of interdisciplinary collaboration. Mark Lemon is a Lecturer at the IERC. He is a sociologist who has worked in community development and has current research interests in the process of environmental perception and the characteristics of resilient communities. He is also keen to explore the contribution that social science can play in providing a clearer understanding of natural processes. Tim Oxley is a Research Officer at the IERC. He has a background in computer science and complex systems modelling and is currently working on the development of various dynamic integrated models of hydrological, biophysical, social and/or economic phenomena. His expertise lies in the integration of spatially and

x

temporally disparate human and natural phenomena and the identification of nonlinear or discontinuous processes. Roger Seaton is Lecturer in Technology Policy at the IERC. He has a background in civil engineering and transport studies and now specialises in research into collaborative technology assessment with industrial and public sector organizations. He has a particular interest in the interface between research and the policy process.

1. POLICY RELEVANT RESEARCH: THE NATURE OF THE PROBLEM Mark Lemon and Roger Seaton

INTRODUCTION Multi-disciplinarity and inter-disciplinarity have become part of the language of environmental education, research and management. Courses move between environmental biology, waste biotechnology, geographic information systems, risk assessment and even into ethics and aesthetics. An increasing acceptance about the complex nature of environmental systems and what is becoming recognised as the ‘seamless web’ between social and natural has accompanied a burgeoning of disciplines and skills under the environmental umbrella. While this is largely desirable two concerns must be expressed. Firstly, multi-disciplinarity can disappear into a generality which constrains the development of single disciplines. Secondly, and central to this book, is the concern that the introduction of more disciplines to the environmental melting pot is considered as the best way to represent the complexity of issues. This is a premature assumption if those issues are inadequately defined and the information requirements inappropriately specified. There is a danger that environmental education can become a catalogue of disciplines, information and techniques rather than a guide to learning how to structure issues or environmental problems (Lemon and Longhurst, 1996). It is this framework for structuring issues which should provide the basis for selecting the requisite contributions from different disciplines. The environment encompasses human and natural interactions and as such encroaches on, and is considered by, a range of disciplines not traditionally linked to natural resource questions e.g. social theory as well as those that would normally be associated with them e.g. the natural sciences. A priority for environmental research and education should be to provide a framework for structuring issues and thereby selecting the contributions to be made by different disciplines rather than moving towards their assimilation under one umbrella. Environmental studies must reflect and make explicit the complexity of environmental issues, including spatial and temporal scale, through the development of a conceptual framework which is not interdisciplinary so much as transdisciplinary. This book is not intended either as a critique of the many attempts at integrated environmental education or to question the fundamental contribution to be made by good teaching and science within single disciplines. Rather, it is an expression of the need to define issues more clearly through the exploration of social and natural interactions and an improved understanding about when, and to whom, those relationships constitute a problem. Therefore, while the insights obtained from ‘traditional’ science about natural processes are of great importance, to be decision or policy relevant they must be related to the socio-economic and cultural environment in which they occur. This has a knock on effect for the way that environmental research is handled, environmental issues managed

2

EXPLORING ENVIRONMENTAL CHANGE USING AN INTEGRATIVE METHOD

and related policies formed. It also provides the rationale for the integrative method which is presented in the ensuing text through a case study of agricultural practice in the Argolid Valley of the Peloponnese, Greece. The book is divided into three sections. Section one opens with a discussion on policy relevant research and integrative method in the context of sustainability. The case study area is then introduced, followed by an introduction to systems thinking and social enquiry which form the generic bases for integrative method as it is interpreted here. The second part of the text examines the agronomic, technological and socio-economic co-evolution of the Argolid Valley over the past fifty years and how this has affected the condition of the natural resource base, in particular ground water. A central feature of this analysis is the variation within the area and an attempt is made to classify this through the development of bio-physical zones and a typology of agricultural decision makers. The combination of these classifications forms the basis for a crop choice framework which highlights the difference between what could be grown (opportunity space) and what is perceived to be possible or likely (decision space). The final section of the book draws upon the information presented in the previous chapters and presents modelled representations of crop choice and the water-salt system in order to move towards the development of an integrated dynamic model. The current methodological, policy and local situations are then discussed and consideration given to the future development of each. Building on the Local It has already been suggested that policy relevant research needs to build upon ‘local’ knowledge and concerns, both in order to specify more appropriate forms of policy intervention, and to have a more informed view about the possible impacts of such intervention. The method described below is an iterative process which is still being explored and developed both in relation to the Argolid and in the Marina Baixa region of Alicante in Spain. It involves interacting with stakeholders1 in the processes of local description, issue specification, impact assessment and interpretation. While these stages are appropriate to the understanding of change in real time they are also pertinent to the generation and interpretation of ‘modelled’ scenarios that can be used to explore a range of possible futures. Local people have been involved extensively in the ‘qualitative’ aspects of the work, however, to date they have been less active in the processes of model specification and the interpretation of the output that is generated. It is a central objective of ongoing work to address this issue and to involve stakeholders at each stage. A second objective concerns the need to establish a more generic, accessible and policy relevant methodology. To this end the project seeks to provide a common method for examining the impacts of land use change in qualitatively different locations experiencing different forms and intensity of socio-economic and bio-physical degradation. It is anticipated that this approach will shine some light on the background to salient issues in the case study locality and in the process of doing so will contribute to a policy relevant framework, that is of value to local stakeholders, for the exploration as opposed to prediction of possible futures under different land use regimes (see McLain and Lee, 1996). For example the Argolid Valley will be seen to have moved towards the subsidised monocropping of citrus fruit over the past forty years and to have experienced a concurrent 1

Stakeholders in the context of this work are taken to be those actors and agencies that affect and are affected by the processes under investigation.

POLICY RELEVANT RESEARCH

3

degradation (salination and depletion) of its natural resources, in particular water. This degradation was of fundamental concern to the area in the early 1990’s when the initial research was undertaken. However, in 1996 and 1997 there has been a marked increase in precipitation levels and the concurrent recharge of aquifers from local springs. These events have led to a shift in local priorities away from resource degradation and towards economic issues relating to the marketing of crops and changes in policy concerning price support and subsidy. While this book addresses the substantive issue of how strategic research and policy formulation can better contribute to the evaluation of sustainable water systems it will attempt to comply with its’ own message of adaptive systems by incorporating these more recent changes into the description and analysis of local options. The overall goal of water policy at the level of the European Union (EU) is one of “sustainable water supply” and this presents considerable operational difficulties for the assessment of spatial, water network (infrastructure) and environmental impacts. The idea of sustainability derives from the desire to pass on to current and future generations the same or an improved quality of life although, given uncertainty, not necessarily the same configuration of opportunities as exist now (Bruntland, 1987). It is thus necessary to avoid reducing that future capability by causing permanent damage to environmental systems in the pursuit of shorter term economic and social viability. The problem for policy at the European level is to help shape that shorter term viability so that it is more congruent with long term sustainability. It is particularly important to consider the diverse circumstances and perceptions that arise from different cultures and traditions across the EU and in the wider European context. This has implications for how the aggregation of behaviour is treated in different locations and in turn what effects and policy options may be appropriate for consideration at the European level. The recipients of policy at the local level will also be responding to a wide variety of local, regional and national influences and the recognition of this diversity is fundamental to the concept of subsidiarity. Indeed the maintenance of diversity of life style is an important feature of a sustainable European Union. It is thus essential that the policy process has access to methods, models and techniques which enable properties and behaviour at each level to be formally linked. At the same time different temporalities need to be considered, particularly the time horizons that are appropriate for strategic policy. The response to changes in water need, and use, may be relatively fast for marginal changes in quality and costs. More fundamental changes such as the introduction of new technologies and infrastructures may lead to structural changes regarding the source and level of demand in urban areas and regions over somewhat longer periods. Such spatial changes will also be influenced by changes in the physical, social, and economic environment alongside the evolution of innovative and spatially different infrastructures etc. The emergence of new inter-urban and regional spatial structures takes place over even longer time periods. In a similar way the physical “environment” is not a passive entity but is driven by a complex set of factors which themselves have spatial and temporal properties. Dispersion and concentration of pollutants in air, water and soil, the rate of transformation of pollutants into other damaging phenomena and so forth all take place over different time scales. Changes in the output of pollution and the immediate affect on a recipient human population are likely to be faster than the speed at which their effects reveal themselves in the ecological system. It is no longer considered sufficient to control, where possible, the output of pollution or use of energy and materials without a better understanding of the how these outputs influence our future ability to function and what the longer term effects are on those affected by them.

4


Some Useful Distinctions A number of the terms and concepts adopted for the work and used in this text need clarification. These are not intended to be taken as definitive statement but rather to serve as a route map through the methodology which has been developed. They are invariably derived from soft systems thinking and have been presented at greater length elsewhere (i.e. Winder and Van der Leeuw, 1997). • Actors are the decision makers (individuals, groups, communities and organisations etc.) in complex situations. • Agencies are those organisations, whether governmental, non-governmental, voluntary, statutory etc. which have any function that influences actors as a consequence of any intended policy implementation at any level. • Policy relevant refers to investigations and research concerning a context and related issues about which policy may need to be formulated by a responsible agency on behalf of a wider group of organisations, communities or citizens. • Issue relevant refers to specific contexts in which there are symptoms which the embedded actors perceive as a “problem”. • Decision relevant refers to contexts in which the actors have, either implicitly or explicitly, identified the nature of their problem and the choices and options and wish to further their understanding of the options and consequences. • Decision space refers to the range and nature of options considered by the actor(s) to be relevant and potentially achievable. • Opportunity space refers to the number and nature of all the options notionally available to the actor. This will include outcomes of decisions which are not perceived by the actor or cannot be considered viable in terms of their ability to access them. • Policy formulating process refers to the qualitative, political, administrative and scientific interactions by which policy is derived. • Policy instruments refer to the various types of intervention (economic, infrastructural, educational etc.) which can be used to enact policy. • Policy delivery refers to the actions, mechanisms and management required for a policy to be enacted by agencies in order to impinge on the final recipients of policy. Often there is confusion between policy and policy instruments since within a hierarchy policy instruments at a higher level only reveal themselves as policy at the level down. In the case of EU policy this often involves a hierarchy of agents from member governments to local government or agencies. Policy delivery can be considered as a logistical problem by which the intended effects of the policy are determined not only by the choice of policy instruments but also by the choice and design of the delivery mechanism. Culturally and economically diverse agencies and recipients are likely to respond differently as are those agencies responsible for policy delivery. This suggests that in order for policy to achieve equivalent, but not the same, effect it may the policy instruments, and delivery mechanisms employed, to have sufficient variety to match the range of contexts to which it is delivered. • Decision-making process refers to the sequence in which knowledge is used and the procedures by which decisions and subsequent acts and activities are derived and evaluated. Both this and the policy formulation process may be implicit rather than explicit and under some circumstances the actors may not consider that they actually agreed a policy or a decision.


5

• Decision making is itself a complex issue but can usefully be thought of as involving a number of attributes of the elements (decision space) relevant to the actor(s) involved. Thus a decision takes place in an attribute space specific to the actor. Where there are very similar decision issues and very similar decision spaces then the relevant attributes of one actor and another in the same cultural context will be similar and the probabilities of individual outcomes can be aggregated. More commonly, each person will operate in an attribute space which has some common attributes with some people or organisations but which may have others which are different. When the attribute spaces are sufficiently different as to result in systematically different outcomes then diversity is apparent. When the decisions concern the same attribute space but different combinations of intensity of each attribute then variety is apparent. It should be noted, however that what is considered a decision to an outsider may not be considered as a “decision event” by the actor(s) involved. (The notion of “a decision” may itself be a scientific or cultural construct). Similarly the decision space of a policy recipient may be very different to that of the policy which attempts to obtain change in the recipient’s behaviour. Decisions and their pertinent attribute spaces are not just hierarchically juxtaposed but are also nested in the sense that on some occasions decisions by, say individuals, form the apex of a hierarchy while in other situations they are the object of decisions. Decisions at any level interact not only with a given issue but with other issues at different levels of phenomena. It has been suggested that delay is often apparent between ecological distress and stakeholder perception of an ‘issue’. This proposition can be extended further to include the delay between that ecological distress and the ‘issue’ as it is interpreted by non-local institutions or agencies. This suggests that issues have their own emergent characteristics which are grounded in perception and decision making, bio-physical processes and institutional responses. The concepts introduced above are intended to provide a framework within which such propositions can be explored. PARADIGMS, PERCEPTION AND PROCESS: SCIENCE AND DECISION MAKING In their reflections on the need for a more useful framework for analysing policy problems, several studies have concluded that the conventional scientific model is not appropriate (Dunn, 1981;

Figure 1–1 The ‘rational’ view of science and decision issues.

Cleveland, 1988). A major reason for this is the discipline specific approach which it promotes and tends to adopt. In his review of paradigms in policy making, Daneke (1989) states that:

6


‘A significant paradigmatic shift is severely hampered by extreme disciplinary isolationism on the one hand and the pervasive imperialism of neo-classical economics on the other.’ Referencing the pioneering work of Holling (1978), Daneke calls for ‘modest conceptual developments, seeking new applications for notions such as resiliency, adaptive learning, and other institutional dynamics.’ In other words there is a role for research which focuses on the “process” of change and which provides insights into how policy may guide it. The traditional, rational view of science research about decision issues is shown in Figure 1–1. This approach reduces complexity and seeks to identify simple causal relations. The adoption of such a paradigm is invariably considered necessary because it provides an appropriate way to develop an understanding of classes of phenomena about which scientific methods were developed and about which they are efficient generators of new knowledge. It can then be argued that policy should focus on those causes. In the complex, messy real world (Checkland, 1981) the consequences of such a view often result in “unintended” consequences which in turn become a new set of “unanticipated” decision-issues (Figure 1–2). Invariably the structures and procedures in place to manage change are based upon end state planning and an inability to respond, or even recognise, the complexity of the process. This ‘closing’ of the system in order to match it to formal management structures and practice invariably fails to account for the informal responses and interactions that form the basis of self organisation. In consequence there is often a poor level of congrence between this formal representation and the issues as they are interpreted at a local level (Lemon and Naeem, 1990). The conceptual model shown in Figure 1–3 assumes that a “decision-issue” has been correctly identified. This presupposes firstly that symptoms of a “problem” have been appropriately linked to a single decision issue, that there is a single homogeneous audience for change, (the problem owners) and that the problem setting can be well bounded (i.e. there is no interaction between it and anything else). Secondly, there is a problem concerning the extent to which it is assumed that in future the causal factors and the problem setting will remain the same. Such an approach is based upon the false assumption that the future is forecastable. Although considerable progress has been made in handling the concept of risk, the notions of uncertainty and surprise resist formal analytic techniques (Jeffrey and Seaton, 1995). If it is inherently impossible to forecast the particular configuration of a situation in the future then it may be more appropriate to emphasise those attributes that are connected with the capacity to adapt and change, on an ongoing basis, such as resilience and flexibility. These limitations of the traditional deterministic approach have quite serious implications for strategic decision-making although there are other types of decisionissue for which it may be appropriate. For instance, decisions about physical infrastructures still have to be handled with regard to the problems of engineering design and implementation. Thus end-state planning, while inappropriate for strategic decision issues, cannot be avoided in others that are ‘project’ oriented. Much depends on how artefacts and devices are embedded in ever changing organisations and communities. In consequence, there is a tension between technological solutions to problems with long lives and the rate of change of the context or location to which they are applied. Figure 1–3 therefore conveys the traditional role of the technical analyst with respect to decisionmaking, the recipients of policy and the decisions themselves. Even in contemporary research this


7

Figure 1–2 Congruence model for planned/unplanned change.

Figure 1–3 Technical perspective on decision making.

model is surprisingly common. Scientists and engineers are particularly vulnerable to its adoption when they apply themselves to policy issues. The most striking feature of it is the notion of “decision-maker” that is commonly used. In practice the vast majority of strategic decision issues involve a complex decision-making process which includes a diversity of information, views and participants. The idea of decision making as a complex, human centred process is common in politics, management and social policy, all of which may use formal intellectual devices with different traditions to natural science. Increasingly there is an emphasis on the role of social disciplines in formulating the appropriate decision context to which science and technology knowledge has to be applied (see Newby, 1992 and Chapter four). There are further limitations in the application of a traditional science model to decision making. Firstly, it confuses accountability with agenda setting. In other words it sees the role of the analyst as providing information to the “decision-maker” on an issue that has been determined by that “decision-maker”. This makes the questionable assumption that the agenda for change of the recipients of decisions and policy is fully comprehended by the decision-maker. (Lemon, Hart and Seaton, 1992). In other words the relationship between the decision-making process and the recipients of policy and decision making is assumed to be about the selection of options and not about option generation. In practice the amount of knowledge at the disposal of the policy

8


formulating process about the priorities for change and the likely response of a recipient population to a specific policy or policy instrument is often very limited. A second limitation of the traditional model is the nature of the relationship between the researcher or analyst and the recipient population. This is shown as a broken arrow (Figure 1–3) to denote the lack of interaction between them. The recipient population is usually seen as a set of objects to be observed and analysed. Where they are questioned it is invariably about the researcher’s agenda without adequate regard being paid to it’s relevance for the respondents. The problem is that the analyst takes no responsibility for enquiring into those diverse agendas and decision spaces. The consequence is that the relevance of the nominated issue to people on the receiving end of policy and strategic change is never known (Lemon and Naeem, 1990). Thus one essential characteristic of policy relevant research is that attention is paid to the agenda for change and the receptivity of the recipients of change (Seaton and Cordey-Hayes, 1993). Receptivity in this context can be seen as the ability to assimilate change. The weaknesses of the traditional approach have to some extent been recognised in the social disciplines through the development of techniques and mechanisms for the elicitation from populations about their perceptions of relevant “problems” and the prioritisation of decision issues. A strong complementarity between those developments and social anthropology has been revealed during this project (Green and Lemon, 1996), in part because of the contribution social anthropology can make to an understanding of social and cultural diversity and the influence that social and cultural factors have on the decision-space of policy recipients. One of the problems of much current policy formulation in the technological and environmental fields is that it not only affects the target recipients, sometimes in an unexpected way, but that it also affects other unanticipated groups because of physical and social linkages that can only be identified through interaction with the population. To formulate a strategy and policy without regard to the likely response of the recipients is unhelpful but what is also required is the formal recognition of the diversity of agendas for change and the range of priorities for different individuals and groups. This calls for the adoption of appropriate techniques of enquiry into the agenda of needs and requirements and how these change over time. In summary the traditional approach of the analyst and the related view of strategic decisionmaking has severe limitations when applied to decision issues that involve interactions between the social, technological and natural worlds. The approach that is presented in this text is a consequence of the increasingly evident failure of any single epistemological position or intellectual discipline to provide both the diversity of knowledge, and the integration needed, to formally link research disciplines together and to extend this formal linkage to a policy and decision making setting. While we do not claim to have progressed far along the path, we do feel that there is a need to look beyond the addition of different disciplines and to consider how best to harness their synergistic potential. MOVING BEYOND MULTI-DISCIPLINARITY What is presented below is an attempt to show how existing intellectual resources can be exploited to explore decision issues without resorting to such devices as “meta-disciplines” or “metamethods”. It is best seen as the application of a technological perspective to research knowledge in that it attempts to exploit knowledge generated from traditional disciplines and to show how such knowledge can be integrated in a way which makes it accessible to policy and decision making. It


9

does not depend in itself on a new theoretical framework but adapts conceptual devices and models from established and developing fields of research. The distinction is, therefore, between the development of methods for knowledge exploitation to support decision-making and the generation of new knowledge through theoretically based research. Since such a decision relevant approach aims to exploit the knowledge in existing disciplines, it must be judged, as with all methods, by the usefulness of its application in the context of policy formulation and strategic decision-making. Thus the appropriate criterion is the extent to which such an approach provides “additional” knowledge and information for decision-making that is not available through the pursuit of a single epistemological position, nor indeed from the ad-hoc application of a number of these to the same issue in an unconnected way. Experience suggests that in such multi-disciplinary research the need for post hoc integration is seldom attained. This method is based upon the view that the world about which decisions are formulated and acted upon are, as noted earlier, dynamic, uncertain, messy and complex. Dynamic: Not only that there exist many interactions but that these are changing over time. Complex: While a component or sub-system affects another, it too is also affected by those it affects (feedback). Uncertain: It is not possible to forecast future states because it is certain that unpredictable events will occur which will affect any assumed situation so radically that a further, unknowable vector will be followed. Messy: Checkland (1981) asserts, that in reality there are no such things as “systems” (although conceptualising certain aspects of the world as though they were systems is certainly useful), thus messy denotes conflicting goals, values, behaviour and irrationality (at least in terms of any particular epistemological position). In these situations, which are the ones that policy and decision making inevitably address, no one method of enquiry can expose the knowledge needed. On the other hand the knowledge required is considerable. Clearly just making bigger “models” of greater complexity is unhelpful as is more measurement without regard to the significance of that data to decision issues. It is useful to consider the nature of the contribution that formal research disciplines can make. Each discipline or system of thought that is relevant to the class of decision issues can be considered in two ways. 1. It consists of internal debates and enquiries which facilitate its evolution and which can provide knowledge about a certain class of phenomena and issues. 2. Thus the discipline and its sub-sets have their own internal dynamic, shared conceptual devices and agenda. Part of the Argolid project, which is reported below, has undertaken an initial investigation into the intellectual position of scientists concerning the “problem” as they perceive it. However, embedded in that knowledge is some component which could be useful in a policy context. The difficulty is that the relevant knowledge seldom addresses directly the policy issues from a decision making point of view. It is articulated in a way that is appropriate to that discipline and not one that is relevant to policy formulation and decision making processes or to the recipients of those processes. The practical question arising out of this is how to arrange circumstances so as to obtain “emergent” knowledge, i.e. knowledge over and above that of a number of disciplines. Experience of multi-disciplinary research shows that not only does it often lead to conflict (resulting from attempts to redefine the issue to one which is tractable within one discipline as opposed to

10


another) but that we are left with a number of qualitatively different reports and insights. While it is apparent that no single one of them is sufficient we must give some consideration to the development of mechanisms that can combine them in such a way that we end up with more than the sum of their particular data sets? PATHWAYS OF CHANGE Policy relevant method is therefore grounded in the identification of salient issues at the local level. This does not assume homogeneous responses, indeed it recognises that the complicated picture that may emerge is invariably the subject of political arbitration or prioritisation. Similarly the local need not represent only the smallest unit but the representative voice of stakeholders at their constituent levels i.e. the village council, the farmer, the agricultural co-operative. Representatives of these groups may have very different perspectives and some individuals are likely to operate in multiple capacities which on occasions appear contradictory. By exploring these perspectives an improved, but often less clear, description can be obtained about how issues are defined, the processes that are seen to impact upon them and those that are affected by them. These perceived ‘pathways’ of change do not necessarily coincide with the political or academic agendas relating to pre-defined issues and as such can provide useful insights into the unanticipated consequences of planned change. It has been argued that individuals often assimilate process in a more complicated manner than can be understood simply by focusing upon the changes in attributes defined by technical agendas.2 It is of primary importance to establish whether those attributes, and the issues to which they refer, are of relevance to the stakeholders concerned; indeed as part of this process it may also be necessary to reappraise who those stakeholders are. At the heart of policy relevant method are a set of procedures for establishing how change processes are perceived and how decisions might be influenced by that perception. It is this which provides the basis for local description (Murdoch, 1995) and which identifies attributes and relationships that may be overlooked by more structured ‘participation’ exercises. By pursuing this description it is apparent that multiple possibilities for intervention develop and that the scope for future uncertainty is accepted rather than obscured by the procedures of technical simplification. This requirement to ‘ground’ policy relevant method must reflect and make explicit the complex perceived paths which define environmental issues. These include the spatial (geography and organisation) and temporal (duration and tempo) components of substantive processes and the agencies or stakeholders involved. For example the provisional analysis of the Argolid data has considered a range of change trajectories and the relationship between agencies and the scale and location of that interaction. While the ‘projects’ envisaged within ‘closed system’ approaches (Figure 1–4) vary considerably they can only be usefully evaluated in the context of how they are likely to impact upon the wider system and how those impacts are going to be interpreted by the local population. The engineering approach, which was adopted for the building of the Anavalos canal system for distributing spring water set in motion processes which ran counter to the original intention of the project (Figure 1–5). The building of the canal did provide irrigation water to large parts of the central Argolid Valley which had been particularly vulnerable to salination resulting from sea water intrusion. It followed a course through the central plain which meant that those who farmed the areas on the periphery did not have access to the transported water. They were also subject to less salination but had suffered


11

Figure 1–4 Technical (closed system) perspectives on the issue of degraded water.

Figure 1–5 Open systems interpretation of new canal infrastructure.

resource depletion, much of which was accredited to the activities of the predominantly monocultural farming activity in the centre. Furthermore, the peripheral farmers felt that they were ‘authentic’, full time farmers whereas those in the centre were often perceived to be ‘inauthentic’, farming labour extensive crops because they had primary occupations and used their earnings from farming as a substantial income supplement. This situation was perceived to be caused or certainly compounded by the European Union’s price support for citrus crops which are labour extensive and heavy water users. This brief sketch is introduced to highlight the limitations of adopting restricted ‘technical’ perspectives. In so doing it makes clear that the objectives of a scheme may well be met i.e. through the transfer of better quality irrigation water and the possibility of improved aquifer stocks in the short term. However, when the quality of those stocks make them accessible again through bore-

2

This may not always be matched by the ability to articulate process (see Giddens on levels of discursive consciousness, 1990).

12


holes it is likely that they will become so because little cost is incurred in the central areas where the water is close to the surface and at present the water provided by the infrastructure is charged for. The degradation spiral again becomes a real possibility. Similarly the existence of price support for water heavy crops has reinforced the need for additional technology, however, there is a current move away from price support. Of equal importance has been the perceived inequity articulated by those on the periphery of the valley who felt, during the period of low rainfall prior to 1995, that they are not only seeing their natural resources decline, in large part due to activities beyond their control, but that they are were also losing out on the opportunity of relatively ‘easy money’. With the current change in emphasis away from price support these farmers are now more confident of their ability to compete in a free market situation. By opening up a system of interest it becomes clear that the noise which emerges is not extraneous but is the basis of the uncertainty that has to be managed. This is a fundamental condition of policy relevant research and central to it is the variation with which processes of environmental change are perceived and the spatial-temporal scales over which they occur. Human systems and their bio-physical environment are, therefore, complex and dynamic; changing and reproducing through time and across space. Each element of a change is the cause and effect of other processes, and as such cannot be measured using a baseline state and a subsequent movement away from that state (as in a closed system). Hence, the process of change can be viewed as a series of concurrent pathways within which interlinking themes or projects can be identified (Figure 1–6). These themes will appear linear in nature and, as has been suggested are often treated as such with the adoption of technical perspectives. However, it can be argued that the interaction of the non-linear elements in the process are often a more influential, although more complex indicator, of the possible outcome of that process. Central to any process oriented approach which would seek to investigate perceptions of change is the proposition that individuals have a cognisance of process as well as of attributes. Any change in state must be seen in the context of a synthesis of other states and not merely as a trade off between one condition and another. It is suggested that individuals continually undertake this synthesis by updating the information they receive from external stimuli and the response to their own actions.

Figure 1–6 Multiple pathways and issues.


13

In summary, individuals seek to cognitively map their environment as a series of social interactions that take place within, and interact with, their bio-physical surroundings (Urry, 1987). The interaction of the natural, built and social environments form an idiosyncratic ‘sense of place’ that is continually evolving alongside, and in response to, local and non-local restructuring and reproduction (Lemon, Green and Filippucci, 1997). The approach to eliciting and mapping personal evaluations of change which is adopted in the case studies is rooted in an evolutionary and multidimensional interpretation of these phenomena in which one cited occurrence or process can at any one time be either cause, change or effect depending upon the context in which it is observed (i.e. simultaneously and in different processes, or at different stages in the same process). It is the diversity inherent within these pathways that conveys the variety in an open system, and in turn is responsible for the unanticipated consequences and uncertainty that end state planning has traditionally had problems with. TIME, SPACE AND AGENCY If we continue with the Argolid example of the canal (Figure 1–5) it becomes apparent that there are multiple agencies (i.e. farmers, engineers, hydrologists, agronomists, central government and the European Union) and processes (i.e. canal building, irrigation cycles, soil-water-salt cycles, economic cycles) operating at different hierarchical levels and over different time scales. These need not coincide with the anticipated progress and impact of the infrastructure. In abstract terms time can be seen as the duration of a process and the variation in tempo whereas space is interpreted in terms of geography and organisational scale (Figure 1–7). More in depth analysis using this framework will be reported later, however, for the time being it is worth re-iterating that an improved understanding of these pathways of change can inform about the range of options for intervention. Central to this understanding is the need to identify the appropriate level of investigation and perhaps more importantly the linkages between levels (see Hofstede, 1995). The pursuit of a multilevel analysis will inevitably cross disciplinary boundaries (i.e. psychology-sociology-policy and economics) and as such highlights the importance of multi-method approaches.

Figure 1–7 Organisational/geographical scale relating to irrigation.

14


It has been argued that it is only through an improved understanding of the context in which an environmental change is perceived that the issues relating to it, in terms of impact, can be anticipated. The systemic picture which emerges from this (i.e. agencies and interactions) forms the basis for specifying an issue, the information requirements associated with that specification and thereby an improved basis for anticipating possible futures (Figure 1–8). One final point pertaining to the development of policy relevant method is the need for such a method to be transferable. The framework introduced above is generic however some consideration needs to be given to the transfer of insights from one location to another. If such a comparative framework is not available then the relevance of a piece of work may well be restricted to single locations and time periods. This raises a number of fundamental problems about the selection of case study examples as the basis for examining more far reaching processes such as desertification and has been discussed at length elsewhere (Lemon, Green and Filippucci, 1997). CONCLUSION: SUPPORTING DIAGNOSIS Figure 1–8 indicates a number of steps that are not disciplinary based but which are considered central to environmental diagnosis and as such should be incorporated into policy relevant research. This translates into procedures which require an appreciation of systemic thinking, social enquiry and conceptual and computer based knowledge or models. These skills sets will be introduced in chapters four and five, they will also be considered as the basis of the integrative method developed for the Argolid case study in particular but with the intention to provide a framework that is transferable to other locations. Each stage represented in Figure 1–8 is based upon the need for interaction—to a greater or lesser extent—with stakeholders. These can be summarised as follows: • the ‘system of interest’ which defines an issue should be articulated by the population(s) concerned (issue specification). • the mapping of this system should identify the range of information, and techniques required, to move towards problem diagnosis (issue representation and information specification). • the future options that emerge from that diagnosis need to be generated and explored (scenario generation and issue (re)specification). In conclusion the approach adopted to policy and decision relevant research draws upon a range of

Figure 1–8 Iterative and diagnostic framework.


15

important ideas, from a number of fields of research. Firstly “pathways” are meant as the vectors of successive desirable change. These vectors can be seen in terms of a number of attributes which are associated with long term survival and evolution. Concepts from various branches of evolutionary theory such as resilience, adaptability and diversity are important and contribute to ideas about sustainability. There is also the difficult question of time scale. Access to future opportunities is an important attribute of survival but sometimes a short term reduction in opportunities may be a condition for an increase in future options. There is, therefore, some sort of balance to be struck between short-termism and long-termism. It is not possible to optimise with respect to the long term since it is certain that any criteria for optimisation depend on a forecastable future. Policy has to proceed towards the future by focusing on the risk and consequences of being wrong and the development of survival attributes in the face of change. This raises two fundamental questions for research relating to sustainability which will be considered in the context of integrative method in the next chapter: • How can research be undertaken which aids in the planning of desirable and viable futures? • How can research be made more relevant and accessible to policy formulation and implementation?

16

2. TOWARDS AN INTEGRATIVE METHOD Mark Lemon and Roger Seaton

SUSTAINABILITY Sustainability is a complex concept which is generally seen as human centred, long-term and involving interaction with natural systems. Many attempts have been made to define the term in the context of global issues (Brown et al., 1987; Liverman et al., 1988; Pearce and Turner, 1990), although O’Riordan warns of the dangers, describing its definition as an “exploration into a tangled conceptual jungle where watchful eyes lurk at every bend” (O’Riordan, 1985). However, despite the difficulty of developing an operational definition there appears to be global consensus about the need for human actions to be sustainable (Earth Summit June 1992; Bruntland report, 1987). Despite, or perhaps because of this difficulty the concept has been the subject of considerable research and debate. The conceptual foundation on which many definitions of sustainability are based differ considerably and are often dependent upon the underlying conceptual models derived from the traditions of particular disciplines. Murdoch (1993) suggests that at a conceptual level sustainability clearly poses a challenge to the maintenance of traditional disciplinary boundaries both within social science and also between social and natural sciences. Human intervention has changed the evolutionary patterns of ecosystems, and continues to do so. The sustainability of agriculture is linked with this process and is closely associated with the dynamics of ecological and socio-economic change. Harwood (1989) takes account of this continual change and defines sustainable agriculture as a system that “can evolve indefinitely toward greater human utility, greater efficiency of resource use and a balance with the environment which is favourable to humans and most other species” There is widespread concern as to whether the farming methods used today are reducing the options for future generations to utilise the land for productive purposes and as such are having a detrimental impact on the ability to farm and produce food in the future. This does not necessarily mean that they will not be able to produce food at any given point in time, only that their options for selecting different futures at a given instant may be reduced. Thus it is possible that the adaptability of farming systems is being diminished. Sustainability in this context can be viewed as the maintenance of the adaptive capacity of farming systems. This adaptability is alluded to by

18


Pearce and Turner (1990) who suggest that sustainable development within a given “vector” should allow development characteristics, (which are open to ethical debate), to be non-decreasing over-time. This highlights an interesting position which suggests that change must be economically viable if it is to be successful. Although this may be the case in the short-term it may not be so over longer periods of time because it impinges on the ecological or natural functioning of the agroecosystem. In this sense increasingly sustainable land based production may be achieved by the identification and discouragement of innovations which are economically viable in the short-term but lead to the removal, or decrease, of agricultural options in the long-term. For instance, investigating sustainability within the Great Lakes Basin, Slocombe (1990) states that the “monitoring of progress toward sustainability depends on identification of system characteristics that either support or decrease sustainability.” Whether or not change is deemed to be sustainable will depend to a large degree on the current state of the system. The adoption of a certain cropping innovation in one situation, (with prevailing climate, soil type, and socio-economic conditions etc.), may be deemed sustainable, whereas in another situation it may well be decreasing the degree of sustainability of that system. This suggests that an agricultural change can only be identified as sustainable when the present state of the system into which it is to be introduced is known. This reduces considerably the generalisations that can be made from one site or region to another and as such complicates decision making processes. Decision making is further complicated because the results of research in a variety of disciplines cannot be easily interpreted at the decision making level, and therefore cannot sufficiently inform the change process. Therefore, although the introduction of the term sustainable into planning is not new,1 much conventional disciplinary research has emphasised end states rather than the transition processes or pathways that emerge in pursuit of them and in so doing can dramatically affect their relevance. ADAPTIVE CAPACITY AND RESILIENCE There is now a well established body of knowledge in both the planning and other socio-economic literature which suggests that concepts of resilience, diversity, adaptability and flexibility provide a useful framework for problem analysis. Important contributions in this field have come from Tintner (1941), Weaver (1948), Alchian (1950), Simon (1962), Winter (1964), Jantsch (1967), Prigogine (1984, 1985), and Mannermaa (1991). The diffusion of these ideas has been spurred by contemporary critiques of mechanistic approaches to problem solving. Both Milliken (1987) and Allaire and Firsirotu (1989), writing in the context of management approaches to planning and strategy, identify the ‘predict and prepare’ philosophy as suspect and propose flexibility as one element of an alternative approach to strategic planning aimed at a “responsive decision system capable of rapid and effective learning and adaption” Ackoff (1983).

TOWARDS AN INTEGRATIVE METHOD

19

Indeed, the use of biological analogies can now be found in literature spanning many disciplines, e.g. Organisation Theory, (Beer, 1966), Complex Ecological Systems (Allen, 1992; Walters, 1986), Economics (Nelson and Winter, 1982), and Development Planning (Sachdeva, 1984). One effect of the transfer of biological and evolutionary analogies to non-biological related disciplines is a broadening of the research agenda to include previously peripheral issues such as long term system survival and the influence of diversity and adaptability on system performance. In particular, diversity and adaptability can be seen as providing options for change in uncertain or unstable operating environments, resulting in more resilient systems. In the light of this policies which focus on diversification as a means for achieving resilience need to pay attention to the following issues: • The relative importance of system level behaviour over the behaviour of constituent elements of the system. • The possibility of reformulating criteria to promote characteristics which might enhance the performance of the system as a whole. (Hierarchical aspects of system performance are particularly important here.) • Altered criteria for system aims such as survival and the continuation of minimum performance levels. • The possibility that there may be more than one path via which a desired system state may be achieved. • The relationship between the benefits of a diversified base and the nature and levels of turbulence in the operating environment. • The relative importance of human centred factors in achieving resilience by way of flexibility and adaptivity (Lemon and Scamans, 1997). AGRICULTURE: SUSTAINABLE PATHWAYS AND INTEGRATIVE METHODS Spedding (1991) argues that although the future is inherently uncertain our inability to predict does not release us of our responsibility for thinking about it. There are three important issues in agricultural research which need to be addressed in order for us to assume this responsibility more effectively. These refer to 1. The broadening and integrating of research so that more relevant information can be made available to the decision making process. 2. The operationalisation and maintenance of flexibility and adaptability and identification of the types of policy that enable increased options in the future. 3. The identification of longer term attributes, as opposed to narrowly defined states, towards which it would be desirable for (agricultural) systems to evolve. The provision of information for the decision making process requires methodologies for synthesising research and data from a range of disciplines. No attempt is made to suggest that this is the only method of carrying out research toward a sustainable systems framework, however Newby 1

For instance within the fishing industry the notion of sustainable catches have been pursued for several decades, see Allen and McGlade, 1989.

20


(1992) argues that social science is integral and not merely marginal in the understanding of how scientific excellence and technological innovation may lead to economic and social well being. The integrative research approach adopted for this research offers a single configuration of research activities that demonstrates linkages between several disciplines, (horizontal integration) and provides the vertical integration between science in its broadest sense and policy. Therefore, research interfaces enable the presentation and interpretation of information derived from scientific fieldwork at various decision making levels. The concept of sustainability is driving an emergent philosophy in which there is an awareness today of the requirements of future generations. This inevitably means that planning within a sustainable systems framework requires some estimation of the needs of future generations. However, as has already been argued, there are clearly dangers of predicting the future and then planning for it (Ackoff, 1983). Those systems which are seen as sustainable today may be undesirable a decade into the future. This suggests that the capacity and ability to respond to continually changing ‘desirable goals’ may be more important than aiming for set end states at a particular moment in time. In the words of Fresco and Kroonenburg (1992), “in order to be sustainable, land use must display a dynamic response to changing ecological and socio-economic conditions.” Future planning may therefore need to change qualitatively from attempts to achieve measurable end states toward goals that reflect the attributes and abilities of technological and natural/human activities. In this situation the maintenance of adaptive capacity within a production system becomes important. For instance in terms of an agricultural system this may mean that soil erosion will reduce the options available for food production at some point in the future. As has already been discussed the human action that causes soil erosion in one area may actually be beneficial in another. Thus a change of cropping on very high quality land may cause some form of degradation, causing a reduction of options in the future, albeit temporary. On low quality land the adoption of this same change of cropping may actually improve soil conditions and increase the options available in the future. This example suggests the need to exercise great caution before attempts are made to recognise one method of production as more sustainable than another. However an assessment framework can be envisaged that draws on the concept of sustainability so long as criteria can be put in place to assess the possible short and long-term repercussions of change. These criteria and a knowledge of the current situation can support a more informed estimate about whether a given change in land use will increase or decrease the options available to produce food in the future? These dynamic pathways should maintain, and hopefully increase the adaptability within a given production system by maintaining trajectories which can fulfill both short term needs, (i.e. be viable), and long term objectives, (i.e. be sustainable). In order to develop along sustainable pathways, changes in land based production need to be assessed and monitored to provide policy makers with information upon which to base their decisions. This requires research approaches which can vertically integrate investigation into both physical and natural systems with social and economic enquiry in order to identify: • Increasingly sustainable pathways • Mechanisms which may be adopted to encourage change along these pathways.


21

In effect, from the policy perspective, this can be seen as a requirement to map viability space and sustainability space onto each other. Viability space can be viewed as a much more malleable entity which can be manipulated via policy and through economic levers. Sustainability space, although being responsive to policy, has a robust core based upon biogeophysical processes which are inherently long-term and often outside the control, at least in the short-term, of human activity. The aim of policy with respect to the concept of sustainability is to ensure that a system develops along pathways which are within both the viability and sustainability space of that system. However the interconnectivity between production systems makes the identification of increasingly sustainable pathways more difficult. For example agricultural and potable water systems are intrinsically linked, yet one system can interact adversely with the other. This example which is central to the Argolid case study suggests that research towards a sustainable systems framework necessarily recognises both vertical and horizontal interactions. INTEGRATIVE METHOD AND INFORMATION INTERFACES The previous section has argued that research towards a sustainable systems framework requires links to be made between scientific, sociological and socio-economic theory, and for information to be presented in a manner which is directly relevant to the policy formulation and decision making process. Such an approach inevitably raises questions about the way in which research should be designed so that information from both the physical/natural systems and the social, economic and cultural systems can be combined. Research frameworks need to demonstrate the connections between a variety of investigative activities and policy. In terms of decision making these connections are as important as the research activities themselves. A number of key factors can therefore be identified as central to integrative method and policy relevant information: • There are significant problems of obtaining data at a sufficient level of definition to be useful but at the same time capable of providing an insight into the variation within the system of study, both through time and across space. This is particularly relevant to the measurements of physical phenomena in the Argolid where data about the natural system is often of a high level of definition but very localised (in time and space). • There is a need to improve our understanding about how different social structures affect and are affected by the natural system and related to this how the variation in agendas, both within and between agencies, can provide an insight into the bases for decision making. • To be policy relevant these two issues must be represented in an interactive format which can inform about the range of possible futures that might emerge from the influence of different policy instruments on social/natural configurations. These three issues provide the underlying structure of the integrative method which has been developed for the Argolid study. Integration in this context is achieved via a series of interfaces (Park, 1993) that link the concept of sustainability to information that is relevant to policy. These interfaces provide one way of linking the conceptual framework introduced in chapter one to the operational activities undertaken for the study and represented in Figure 2–1.

22


Figure 2–1 Research activity and ecological—agrosystem-policy interfaces.

Interface 1. The interpretation of the concept of sustainability at an ecological level Production systems that interact with the natural environment highlight the considerable scope for human intervention in ecological processes. For instance agriculture can significantly effect decomposition process in the soil, and the water industry can impact on the eutrophication of water courses, rivers and seas. A primary task for research within a sustainable systems framework is to identify key variables which are relevant and measurable in time and space. A shortfall in some research is that the methods suggested for monitoring are either too expensive to undertake widely, do not link to fundamental ecological processes, or are not easy to interpret. The value of data obtained over a wider spatial or temporal frame at the expense of precise definition has already been mentioned and is developed in chapter four. Equally, the choice of variables for measurement must be consistent with the system as it is interpreted by the actors within it. Therefore, the ‘desired pathways’ of change (Park, 1993) which are based upon the adaptive requirement for sustainability can only be moved towards if we have information about an appropriate set of starting conditions. This is particularly important in the context of the modelling work carried out during the current project. If we look at Figure 2–1 then it is possible to see that the first interface is oriented towards linking the ‘scientific’ measurement of natural phenomena (i.e. climate, water quality and availability) with the characteristics of the agrosystem process (i.e. cropping systems and production levels). This forms the first stage of the research activity and it is important to note that the orientation of the study towards agricultural production evolved out of an initial set of semi-structured interviews with key actors.


23

Interface 2. The linking of ecological processes to the attitudes and behaviours of the agents of change Having identified the important variables associated with a production system it is necessary to establish the key agents of change and how their activities can influence these variables. With ecological variables the use of computer generated data has meant that the results of experiments can be extrapolated to either predict likely effects of change or to explore ranges of scenarios. Such models can provide useful information for the decision making process. However as the sophistication of these tools increases there is a danger that the mechanisms which they represent become inaccessible to decision makers. There is also a danger that models are used exclusively as output from research, when in effect they may not be the most appropriate form of output, or could be better exploited as one of a suite of tools used to generate policy relevant questions and thereby, information (see chapter five). In situations involving slower ecological processes it may be necessary to identify suitable proxy measures or indicators of change which can be used in combination with model output. However within a sustainable systems framework there is a need for these types of measurement to be relevant to the decision making process, and not for research to be side tracked into proving obscure ecological hypotheses or following pet investigations. The desirable characteristics of such measurements are simplicity, replicability and broad applicability in a form that is understandable to the non-scientist. The second interface therefore builds upon our acquired knowledge of the ecological condition of the system (interface one) and the ongoing activity within that system. It does this by looking at the social-economic and organisational structures that are in place to support it and the flexibility of these to change in particular ways. In other words what are the social-economic and ecological parameters for generating future options. Interface 3. Linking actions and perceptions of the agents of change to policy issues Exploring the effects on ecological processes by the actions of local decision makers is of limited value unless an understanding can be gained about how these agents respond to a variety of policy instruments and mechanisms. A considerable amount of policy research assumes that if a change can be shown to be profitable or financially viable, then it will take place. The farming agenda is known to be complex, (Lemon and Park, 1993), suggesting that more research is required which can evaluate the interconnectivity between system components. Lockeretz (1991) questions “the justification for devoting more attention to one component of a complex process before we can say confidently how strongly the components are linked”. This has implications for the methods used to gather information about the possible effects of policy and how this can be presented. In this context qualitative (cognitive) maps (Swan, 1995) and models generated by semi-structured social enquiry techniques may be useful (see chapter four). As with the first two interfaces the emphasis of the third must be on the presentation of a range of information in a manner which is policy relevant. The third interface develops the range of socio-economic and ecological configurations from the first two interfaces and places these in the context of existing and

24


potential policy instruments. One important aspect of this process is the culture and procedures of the policy implementation and delivery process. This is central to efficient policy diagnosis and requires further investigation, particularly where trans-boundary policies are concerned. The representation of these complex social-economic and ecological configurations alongside the structural and perceptual criteria for decision making has been achieved through the development of a dynamic, integrated model (see chapters 10–12). Of particular importance in the context of this interface and the objectives of this study is the need to convey the information about these configurations in an appropriate manner and then to represent how they may alter under a range of different policy instruments. It is the combination of these two activities that is ‘policy relevant’. A number of characteristics can now be seen as central to policy relevant method. These underpin the need for the selectivity of information and investigative technique which was expressed in the introductory chapter and will vary according to the different perspectives upon an issue i.e. between and within the farming and scientific communities, and to the ecological characteristics of a locality. Alongside this is the need to represent, possibly in a modelling format, how ‘the local’ impacts upon wider scale systems and vice-versa (Newson et al., 1992; Lowrance et al., 1986). Figure 2–2 represents the activities undertaken as part of the integrative method. The initial phase of the method is to establish an overall picture of the system of interest which, with all its accompanying noise, can best inform us about a particular process. Indeed, one function of the diagnostic output is to flag those areas of the system that require further investigation at higher levels of definition, possibly through natural scientific enquiry at the local level—i.e. into the soilsalt interaction in a particular location.

Figure 2–2 Integrative method: the development of the Argolid study.

The data used to establish the system of interest, which constitutes the first phase of the project, can be seen to range from primary data in the form of interviews to documented evidence, often from official sources, and the use of records from public meetings. This ‘triangulation’ (Silverman, 1993) of different types of data is intended to provide an overview of how the issue is defined and the processes and attributes that interact to define it. It is not suggested that the issue has one correct interpretation or definition and as such the navigational metaphor should not be taken too


25

literally as an indicator of ‘position’. The process of defining an issue is not an attempt to reach some form of consensus or an aggregate picture, rather it is intended to establish the variation of views and behaviours which constitute an overall picture and the position of the issue(s) within it. The system of interest needs to be described before we can establish what elements of it require further investigation (Murdoch, 1995). In order to move towards this description it is necessary to interact with the different actors or ‘stakeholders’ in the system and to establish the range of perspectives, however, it is important to look as well as listen and the observational skills of the research team must be related to their abilities to engage in discourse. In the context of issue based research it is unlikely that there will be scope for some of the more time consuming and complicated skills that are required by ethnography. However while the ability to draw upon insights from these disciplines to help interpret both discourse and observation is invaluable (see chapter four) it should not detract from the need to observe economic interactions, production behaviours, social settings, field structures, crop condition etc. over and above those that are elicited through interviews. For example photographs, the local press, council minutes combine within a multi-method approach (Perkins, 1985) to support a more comprehensive picture about the system within which an issue is situated. The description which emerges is, therefore, the basis for more directed research in subsequent phases. Three areas of expertise have been introduced to support this integrative method, these originate from systems thinking, modelling and social enquiry and will be considered alongside the more recognisable and specific forms of enquiry which are based in the natural sciences (i.e. hydrology, soil science) and can be drawn upon as and when required. Firstly however it will be useful to provide a context for the case study which will form the focus for the development and articulation of this method. The next chapter will provide a brief introduction to the Argolid Valley and the creation of a particular landscape that has emerged alongside the rapid increase of irrigated agriculture.

26

3. BACKGROUND TO AGRICULTURE AND DEGRADATION IN THE ARGOLID VALLEY Mark Lemon and Nenia Blatsou

AGRICULTURAL ‘EFFICIENCY’ AND THE MEDITERRANEAN REGION Farming across much of the Mediterranean has evolved in response to the uncertainty of its natural and climatic conditions with small units, part-time farmers and diverse crop patterns. It can be argued that much of European agricultural policy has been based upon criteria of efficiency which have resulted in standardisation and higher unit production. Using the Argolid Valley in Greece as an example it will be seen that the adoption of policies directed at the farmer, the crop and water resources can encourage, and entrench, intensive monocropping of cash crops while exerting considerable strain upon the natural resources of the area. It is maintained that the degradation of natural resources, and the decreased potential for income generation resulting from this, have resulted in a reduced set of options for farmers and the inequitable distribution of those options among them. Agricultural policy has been increasingly directed towards the improved efficiency of farming activities.1 The effect of this in some areas has been to encourage a move towards increased production and standardisation with insufficient consideration given to the impact upon the physical environment (Clunies-Ross and Hildyard, 1992; Manitea-Tsapatsaris, 1986). This has had disastrous effects in areas of the Mediterranean where, over time, farming structures have previously emerged in response to local physical conditions. Paradoxically the failure of some parts of the region to adopt ‘more efficient’ agricultural production systems has been attributed to the continuation of the very characteristics that have historically enabled it to survive without threatening local natural systems. The nature of this change has not just highlighted a perceived ‘backwardness’ in the Mediterranean, it has also led in certain instances to an inequitable distribution of the costs and benefits within its regions, particularly in terms of the ability to respond to declining natural resources. The physical environment of the Mediterranean is inherently more vulnerable than that of the more temperate regions of Northern Europe with considerable variation in the temperature and precipitation levels from year to. This, alongside the geomorphologic and geological variation of the region, has created an environment that will support a wide range of vegetation. However, with the exception of indigenous crops such as olives this variation is not matched by an ability to produce yields at the levels achieved in Northern Europe (Ruiz, 1988). 1For

example see Council Regulations (EEC) Nos. 797/85 and 2328/91 ‘On improving the efficiency of agricultural structures’.

28


Agricultural development in the region has, therefore, been determined by the physical and social landscape in which it operates. Traditionally this has involved high levels of direct consumption and diversity in the crops grown. Agricultural holdings are generally small and divided into a number of parcels with a high percentage of the local work force actively employed in farming, often in a parttime capacity. These characteristics (i.e. multiple job holding, small farms, local markets and diverse low input cropping) have supported an adaptive agriculture capable of responding to the uncertainties of the physical and climatic environment. They are, however, invariably less efficient in terms of the economic criteria that have been central to recent agricultural policy. Indeed where some of these ‘structural deficiencies’ have been overcome (i.e. through the establishment of co-operative organisations) the resulting loss of crop diversity has often been accompanied by the degradation of natural resources. In Greece the failure to meet criteria of economic efficiency has frequently been accredited to the ‘structural weaknesses’ of Greek agriculture rather than the relevance and suitability of the policies introduced for different localities (Green and Lemon, 1996). The failure of policy to account for physical and social difference can force the farming community into a treadmill of intensive practice that ultimately jeopardises the natural resource upon which it depends, and the social context in which it operates. This study will draw upon research carried out in the Argolid Valley in the Peloponnese of Southern Greece. It will provide one example of an agricultural system that appears to be following an unpredictable, unsustainable and inequitable path. The example of the Argolid will be used to argue the need for integrative method and policy relevant research and will be divided into three sections the first of which will introduce the Argolid Valley as the setting for the study. A case will then be made for policy relevant research and the need for an integrative method to support this, particularly in the search for more sustainable, or more correctly, less unsustainable, forms of land-use. The generic skills that have been drawn upon for this integrated approach to environmental diagnostics (social enquiry, systems thinking and modelling) will then be introduced and their complementarity with the disciplinary contributions from both the natural and the social sciences considered. The second section of the book will consist of a more detailed account of the Argolid case study with a particular emphasis upon the links between agricultural production and crop choice, hydrology, technological change and the cultural factors associated with agricultural decisions. The final section will incorporate this crop choice framework into the process of policy relevant modelling which moves some way towards the exploration of how different policy interventions (i.e. water pricing, price support and subsidies) might impact upon the area. The book has been structured in this way so that the essential elements of policy relevant method are presented to the reader through the example of a case study. It is hoped, however, that the detail of the case will not overshadow the basic message which relates to the method and an integrated approach to environmental research and education. THE ARGOLID VALLEY The Argolid Valley is situated in the Argolis region of the North-eastern Peloponnese of Greece. It is accessible, by road and rail, to Athens in approximately two hours and has sea access via the Argolis Bay which also forms its southern boundary. This position has enabled trade access to the capital and northwards to Northern and Eastern Europe. However problems in the former Yugoslavia and currency problems in traditional Eastern European markets alongside an increase in

AGRICULTURE AND DEGRADATION IN THE ARGOLID VALLEY

29

Figure 3–1 Location of the Argolid Valley in Greece.

transit fees for Greek haulage vehicles have all served to limit the attractiveness of this route (EIU, 1992). Movement into the area has also been evident, both historically through invasion,2 and currently through seasonal labour and tourism. The location of the plain, along with its climate and fertile soils,3 has contributed to its place in history, initially as an important centre for ancient Greek civilisations4 and subsequently as the home of Nafplio which became the first capital of Greece in 1833. Over the past forty years the area has experienced a rapid expansion and intensification of agricultural production and is currently witnessing an increasing level of cultural (i.e. visiting archaeological sites) and ‘leisure’ based tourism. (See Figures 3–1 and 3–2.) 2

From prehistoric Indo-European tribes at 3000 BC to the subsequent arrival of Franks, Turks and Venetians. The soils in the foothills are generally stony and of poor irrigability. 4 The historic sites of Mykenes, Tiryns and the Roman town of Argos are all situated in the Plain and the site of Epidavros is twenty kilometres to the south. 3

30


Figure 3–2 Map of the study area (numbers 1–7 are zones adopted for analysis).

The population of about 50,000 is dominated by the town of Argos with 22,000 inhabitants and by Nafplio which is half that size. The remaining population is distributed throughout the plain and foothills in villages of between 500 and 2,000 inhabitants with relatively easy access to the towns and main communication links. Since the 1960’s there has been some migration into the peripheral villages in the foothills by inhabitants from the mountainous areas around the plain. These communities were predominantly made up of shepherds and livestock farmers, and their families who also grew cereals for domestic consumption. Many of these mountain villages are now under populated with a disproportionately large elderly population and inadequate infrastructure. Unlike the depopulation trend in north-west Greece (Epirus) the push engendered by the harshness of the local environment and the difficulty of earning a living was countered by a local pull in the form of the opportunity to obtain land and undertake irrigated agriculture (Green and Lemon, 1996). This opportunity arose with the exploitation of water in the valley and the subsequent transformation of the physical landscape from rain fed farming, scrub and grazing land to one which is dominated by irrigated fruit trees. “The valley was all stubble, some olive trees and pear trees. All the trees you see now have been planted since 1960. Everything started after the bore holes (and the access to water)”.5

Attempts were made to retain the mountain population as far back as 1924 with the introduction of hereditary licences for the cultivation of tobacco. These have been an important source of income for some of the villages above the plain but are now being withdrawn under subsidy. The licenses

5

Interview with farm policeman in Koutsopodi.


31

were employed to grow tobacco on rented land in the valley rather than in the mountainous areas. This was of particular importance with the freeing up of many small strips of land which occurred as a result of the restructuring programmes designed to uproot certain crops in the region.6 As a consequence income was generated away from the villages, thereby reinforcing the movement towards the valley. This provides one example of how local options were available to the mountain population in a way that was not widely available to other areas undergoing depopulation. The response elsewhere was for long distance migration on a far larger scale than has been evident in the Argolid. The area did, however, experience some out-migration, primarily to Germany, Canada and Australia, from the late 1950’s through to the 1970’s. The process has now reversed and returnees have not moved into the towns, as has been the case elsewhere but have often returned to their villages of origin and invariably invested in farming. The proximity of the villages to urban centres and communications networks alongside the economic opportunities available from agriculture have combined to support this process. Therefore even though the local economy is predominantly agricultural, it is not rural in the sense that it is isolated from urban services, markets or employment and investment opportunities. Movement from the mountain villages by the active work force has been to the foothills at the edge of the plain and has been supported by the opportunity to embark upon irrigated agricultural production and the existence of local and European markets for their produce. Some of these peripheral villages have a number of small mountain settlements under their jurisdiction. For example the administrative boundaries of the village of Elliniko includes five or six small communities, however the distance between Elliniko and Argos is only ten kilometres. The head (Proedros) of the village emphasises this relationship when he states that: “I would call it a suburb. It is the balcony of the Argolid. You can see everything, mountains, valley and sea” (Figure 3–3). The statement also indicates the topography of the area which in addition to the Argolis Bay is bordered by the Arachnaios and Arcadic mountains. The area is approximately 1,000 km2 distributed by altitude in the following way: 0–200 (m) 41,900 hectares 200–500 (m) 26,500 hectares>500 (m) 38,600 hectares The central valley has predominantly clay soils whereas the peripheral foothills are characterised by sandy loam, loam and clay loam soils (Figure 3–5). This means that the central area tends to retain water and is vulnerable to water logging with the potential to flood. It also provides a barrier to sea water ingress from the Argolid Gulf but by the same token tends to accrue salts because of a limited potential to leach. The peripheral areas with permeable soils require more water for irrigation and tend to drain into the central valley thereby making them vulnerable to water depletion. The distances between non-urban and urban settlements, including Athens, the topography in and around the plain, and the improving infrastructure have all helped prevent the Argolid Valley becoming marginalised. This is reinforced by a climate which is not perceived to manifest the extremes elsewhere in Greece and is a mix of Mediterranean and Continental European types. Mean temperatures range from 8–10°C in January to 28°C in August, although the extremes can be as low 6

I.e. The subsidised uprooting of Bebekou and Tyrintha apricots in the early 1990’s to prevent the spread of the Sharka virus.

32


Figure 3–3 The Argolid Valley from Elliniko in the South West foothills.

Figure 3–4 The view of the valley from Mykenes in the NE.


33

Figure 3–5 Soil types in the Argolid Valley.

as – 5°C and as high as 45°C. There is a risk of frost between November and March with approximately five days of partial frost close to the sea and twenty five days further into the plain. The frost can destroy the citrus crops and one form of protection against it is to spray with water. This will be seen to be of particular importance when water stocks are low and inadequate control is exercised over the timing and duration of spraying. The mean rainfall of the area is 510mm per year distributed over about ninety days although there appears to be considerable variation in the amount, and distribution, between and through the years (see Figure 3–6 and Table 3–1). Obviously this has implications for any attempts to replenish ground water stocks from springs that are fed by seasonal rainfall. Indeed, with the low rainfall in the late 1990’s considerable concern was expressed by the peripheral regions about the extensive use of groundwater in the main valley. This was a source of considerable antagonism between the centre and the surrounding areas and led to the instigation of remedial activity (i.e. replenishment of the aquifers) and further investigation into resource degradation and in particular ‘desertification’ processes (i.e. the Archaeomedes project discussed in this text). Subsequently, the extreme rainfall which has occurred in the mid 1980’s has resulted in considerable flooding in the central areas, particularly around Nea Kios and Dalamanara. This has led to conflict between residents of these areas and those responsible for aquifer replenishment which was seen as the main cause of the flooding (see chapter 7). The topography of the Argolid provides a backdrop to a landscape which is dominated by apparently lush agricultural vegetation, in particular evergreen orange trees. Over 85% of the wat er used in the area is for agricultural irrigation, with the remaining 15% shared between the industrial and domestic sectors (Association of Argolid Agronomists, 1992). The physical character of the area has therefore been transformed by human intervention and the expansion of irrigated agriculture (Tables 3–2 and 3–3). This is evident on the surface through the

34


Figure 3–6 Rainfall distribution about the mean (510mm/yr) in the Argolid Valley. Source: Meteorological Service. Table 3–1 Rainfall between January-April (mm).

form of vegetation (i.e. fruit trees) and the technology used to support its production, and underground in terms of the impact that such intervention has had on natural resources. The intensification of agriculture, with an accompanying increase in mechanisation and fertiliser and pesticide, has also led to a critical level of degradation in the quality and availability of water, and the progressive degradation of soils. “During the last thirty five years ground water has almost been exhausted, polluted and contaminated, the soils are in danger of degradation and even the quality of spring waters has deteriorated. Argolis today is not just poorer than in the past, but it follows a degradation path that may ultimately lead to Desertification” (AUA, 1993). In the early 1990’s this degradation path was both quantitative, with bore holes in the area reaching 400 metres in depth and with increasing uncertainty about reaching water, and qualitative, with salinisation occurring primarily because of sea water intrusion into the valley as ground water levels dropped. This indicated not only advancing degradation but also the uncertainty attached to it and the increased level of technology necessary to maintain apparent stability. The higher levels of rainfall from 1995 and the replenishment of the aquifers with spring water from the Anavalos springs has since raised the aquifer levels in the valley and reduced the levels of salt in the aquifer. There has, however, also been significant flooding in the central areas and the process of aquifer recharge has been blamed for this by many of the local population (see chapter seven).


35

Table 3–2 Land use in the Argolid Valley.

Table 3–3 Estimated expansion of irrigated agriculture.

IRRIGATED AGRICULTURE: THE CREATION OF A LANDSCAPE The issue of water has been inextricably linked with the Argolid Valley in the north-eastern Peloponnese of Greece since ancient times when the area was characterised as polidipsion (very thirsty) or anydron (lacking water). Greek mythology pays reference to this relationship with the killing, over water, of Hydra by Hercules near the spring of Lerni and the settlement of Mykenes close to freshwater springs being recorded by Homer. Similarly the people of Argos used to drop horses in honour of Poseidon into the springs which emerge into the sea at Anavalos near Kyveri in the west of the area. Older farmers still refer to the local springs as the ‘trimeria’ which occurs when underground acquifers overflow and ground water comes to the surface (Lerni and Kefalari) or when the water directly enters the sea (Anavalos). Records of the area convey a picture of rain-fed farming and grazing with salt marshes along some of the coastal strips. The contemporary visitor to the Argolid Plain, however, will drop from the surrounding foothills into a carpet of lush green vegetation. It is a physical landscape that has been transformed through the access to, and use of, irrigation water. This transformation has occured above ground with the expansion of irrigated agriculture and in its physical form is represented both through the vegetation, with a profusion of citrus trees, and the sight and sounds of the water technologies that have been introduced to support their production. The area is laced with water infrastructure in the form of concrete irrigation canals ranging from six metres in diameter to small channels which skirt individual fields with metal sluice gates for supplying ‘flood’ irrigation to the crops. Elsewhere irrigation pipes snake through the orchards taking water from the crudely built brick bore-hole housings to the sprinkler irrigation systems that are also used in winter to protect the crops against frost. Adjacent to some of these housings can be found the old surface wells that were petrol, and even animal, driven (Figure 3–7). In the coastal plain the level carpet of trees is broken by the presence of air-mixers, a form of inverted wind mill, which are used to protect the crops against frost by disturbing the cold air. Not only do the mixers stand above the orchards, when they are operating in the winter evenings they make a distinctive sound which re-emphasises the dominance of agricultural production in the area.

36


Similarly as the demand for water has increased, and ready access to it has become problematic, the Argolid Valley has become dotted with ‘exploratory’ irrigation rigs. Agricultural production, and in particular citrus production, dominates the senses throughout the year with the powerful scent of the orange blossom and the shrill grinding of chain saws cutting trees for splicing7 in the spring and the muffled accents of the migrant pickers as they walk to work early in the winter mornings.

Figure 3–7 Disused petrol driven water pump.

The visitor to Nafplio between November and March will find the generally peaceful harbour full of boats destined to carry citrus to Eastern Europe and a queue of others waiting in the Bay of Nafplio for their turn. Many of the large warehouses that are closed for the rest of the year become centres of activity for the packing and distribution of citrus crops and the road sides acquire a scattering of small stands selling a range of produce. They also become active with trucks transporting the produce to the docks, to Greek markets or to be buried as excess production. The problems in the former Yugoslavia have, however, greatly reduced the land based transportation of citrus to northern and eastern Europe. Throughout the valley, but especially in the peripheral villages where more diverse cropping takes place, transportation is dominated by small pick-ups trucks. A continuous stream of these can be observed on the Argos-Athens road (approximately

7 The splicing of trees is used to change varieties i.e. from Navalino orange to clementine, invariably in response to changing levels of price support or subsidy to restructure the crops produced. The ‘host’ tree is cut about one foot above the ground and a number of branches placed in notches which form the basis for the graft. A grafted citrus tree may only take two to three years to bear reasonable levels of fruit whereas a new tree can take six years.


37

150kms) in the early morning and late afternoon as farmers take their produce to market in the capital. What this brief description has attempted to convey is, that despite an increasing emphasis being put on tourism, a surface landscape has been constructed around agricultural production. This domination is emphasised by the fact that over 50% of the work force in the area are registered in the census as being employed within the agricultural industry although, as will be seen, this is distorted by the number of part-time farmers. The sub-surface landscape has also changed with the intensification of agriculture, and the accompanying increase in mechanisation. The technologies which have enabled access to water, and have thereby extended the irrigated areas, have now had to be adapted to compensate for increasingly degraded ground water. Therefore, the link between the Argolid and water is now inseperable from agricultural production and related technology. What is seen as the basis of the local socio-economic and physical landscape, the artefacts and processes of irrigated agricultural production, is now seen to hold the key to the future. What appeared as a belief in the ability of technology to determine the course of nature has developed into a resigned acceptance that it offers the only possibility of reversing degradation processes while retaining the ability to generate an income from agricultural production. In the current situation this reversal has been supported by above average precipitation alongside the artificial recharge of the aquifers. Difficulties relating to a reduction in price support are, however, encouraging crop diversification, primarily through increased citrus varieties and thereby a continued reliance upon irrigation and water technologies. In bio-physical terms the extreme rainfall has led to flooding, to the saturation of crops and the reduction in their productive capability, and on occasions to the loss of the trees themselves. If we accept that loss of productivity defines degradation in this area (Green and Lemon, 1996) then both water depletion and excess can be interpreted as degradation. This case study will explore the background to this transformation and in so doing will support the argument that sustainability is not restricted to bio-physical factors but must account for socio-economic and cultural processes. These sub-systems (i.e. biological, physical, technological, socio-economic) and their interactions will be explored in more detail in later chapters, however Table 3–4 provides an overview of the recent changes relating to agriculture that have occured in the study area. The next section of the text will introduce some of the techniques that have underpinned the methodology adopted for this study and are felt to be fundamental to the development of an integrative approach. Chapter four will consider the role of social enquiry and chapter five that played by soft complex systems thinking and modelling.

38


Table 3–4 An historical overview of agriculturally related changes in the Argolid Valley.

4. SOCIAL ENQUIRY AND NATURAL PHENOMENA Mark Lemon

INTRODUCTION The dilemma for anybody attempting to understand how individuals respond to, and manage, change is that people interpret their surroundings in a highly personal manner. This is not only complicated by individual and collective differences but by their dynamic and changing nature. The central question, therefore, relates to the way in which people make sense of their environment and by extension how they develop habits, skills and styles to negotiate day to day life. The response to change occurs through the interaction of the hardware of the physical environment and the software of the cognitive process. People respond to change as they see it. ‘Objective’ technical representations and explanations of the process only provide one element within the cognitive framework. In other words we need to try and understand more fully how people perceive, and interact, with their environment. This is particularly important for assessing the impact of industrial decisions and processes, and the policies relevant to them. Farming decisions are influenced by a complex set of factors which do not differentiate between physical and social processes. Spedding (1991) argues that while this means the future is inherently uncertain it does not “release us of our responsibility for thinking about it”. In order to acquire an improved understanding of this complexity it has been seen to be necessary to move away from clearly defined disciplinary boundaries towards an integrative method which draws upon the expertise of a number of disciplines as required. While the insights obtained from traditional science about natural processes are of great importance, to be decision relevant they must be related to the socio-economic and cultural environment in which they occur (Lemon and Park, 1993; Butzer, 1994). Not only are human and natural processes inseparable, decisions are made upon the perception of their interactions. It is crucial therefore that attempts to represent those processes incorporate the elements that form the basis for decision making and are not restricted to, or even based upon, those that appear relevant to the scientist. Where the social context of physical change is not considered ‘intervention’ will tend to be defined from a technical perspective and subject to a range of unforeseen consequences. Some progress has already been made towards the integration of biophysical and socio-economic agricultural subsystems (see Chapter 2; Park and Seaton, 1995). This tends to have emerged from agricultural systems thinking and a recognition that the decision space within which farmers operate is not restricted to the physical characteristics of the land they manage. By contrast Molnar et al. (1992) are critical of the social science community for devoting too much attention to paradigms,

40


intellectual shifts and ideological studies rather the actual process of agricultural change. They argue that the social sciences will remain largely irrelevant to the evolution of and direction of agricultural science until such time that they attend agricultural science meetings, publish in interdisciplinary journals, or collaborate in joint projects that address actual problems. This monograph conveys the argument that social enquiry can contribute to agricultural science through a range of analytical and conceptual devices (i.e. from systems thinking and ethnography) and through the use of specific tools and techniques for data acquisition (i.e. basic elicitation techniques such as interviews and questionnaires). This contribution can take a number of forms: 1. Through the elicitation of attributes and processes that help to define an issue or system of interest and the investigation of how these are configured to determine the range of options that are perceived by decision makers (Lemon and Park, 1993). 2. By informing about the possible responses to a range of decision options and the thresholds at which those decisions are likely to be made (Potter and Gasson, 1988; Ibery and Bowler, 1993). 3. Through ethnographic insights which provide a cultural context for more directed issue based social enquiry of the sort undertaken for the current study. 4. For the acquisition of data about the use of natural resources and the extent and nature of physical transformations (Lemon et al., 1994). ELICITATION OF ISSUES AND PERCEIVED RESPONSES TO DECISION OPTIONS The subject for issue based research will invariably be determined by the funding agency, albeit often in response to a proposal or tender of some form. This can lead to a deterministic method which assumes that the definition of an issue rests with the ‘client’ and the group undertaking the research and as such invariably adopts a ‘technical perspective’ (Linstone, 1981). This argument has formed the basis for the ‘appropriate technology’ movement1 which in turn has been criticised for adopting such a perspective albeit with the introduction of materials and processes that appear more appropriate to local economies, ecology, knowledge levels etc. In both cases the starting point is with an agreement about what constitutes a problem being reached externally to the population in which it is located. It is therefore, essential to establish the range of interpretations about a phenomenon or process to establish the context within which individual decisions are made. We have already established that ‘decision space’ is a subset of the opportunity space which determines the emergent paths of complex systems (Clark et al., 1995). It has also been argued that what is required in the initial phases of issue based research is a description of the diversity of local agendas and not an aggregation in response to predetermined questions. Therefore, the elicitation of process must be distinguished from techniques that are designed to establish anticipated behavioural response and the thresholds at which those behaviours might occur.

1

See the debate between Stewart (1987) and Eckaus (1987).

SOCIAL ENQUIRY AND NATURAL PHENOMENA

41

Cognitive Mapping Individual interpretations and perceptions are not only invariable unique they are also dynamic and as such is extremely difficult to represent. Lemon and Jeffrey (1997) have argued that individuals can instil a sense of order about the complex processes of which they are a part through the use of cognitive maps. These maps can take the form of spatial temporal pathways which are reproduced through interview and are analogous to the emergence or irreversibility within complex systems. For example Figure 4–1 provides an example of a cognitive map derived from an interview with a state agronomist discussing water issues in the Argolid. He identifies a range of themes that might warrent further consideration i.e. the culture of public administration, attitudes to long term planning. Such maps, and the interviews upon which they are based, are however subject to the limitations of discursive consciousness or the ability and willingness of individuals to articulate what they know or believe (Giddens, 1991). It must be borne in mind that the failure to elicit responses of a particular kind is a comment on the adopted technique and not the world it is employed to look at. Too often the blame for inconsistency or failure is levelled at the subject matter rather than the methodology. Much of the work which has used and developed cognitive mapping techniques has been relatively formalised within the context of decision support and problem solving (Eden et al., 1992; Madu and Jacob, 1991; Swan, 1995) and based in psychology (see Kelly’s Personal Construct Theory, Adams-Weber (1979)). They are schematic representations of individual or collective cognition about a process and convey relationships in a number of ways i.e. proximity and or causality between concepts. In the Argolid study cognitive maps were developed out of a first phase of semi-structured interviews as a way of conveying how individuals articulate the systemic characteristics of agricultural production and water use. The maps were developed as a set of related statements which gave some indication of perceived causality but more importantly identified the range of factors and processes that interacted with agricultural production and water use. Attitudes and Behaviour No attempt was made to anticipate behaviours from this mapping process. Rather the intention of the exercise was to contextualise the processes through the ‘eyes’ of individual actors although this did provide the basis for more subsequent, more structured enquiry which attempted to establish the range of responses to the perceived risks facing agricultural production. The relationship between attitudes and behaviour is one of fundamental concern to the social sciences particularly when the methodology does not incorporate data about ‘actual’ behavioural change and is based upon relatively few attributes, often determined by the researcher(s). Miller (1980) expresses this concern through a criticism of the weight that can be put upon research that emphasises the use of scales as a way of relating attitudes to behaviour.2 Although talk is not always cheap, it usually involves fewer costs than incarceration or selfimmolation. Yet when the talk consists of Likert or Semantic Differential Scales this commonsense fact apparently escapes many researchers.

2

I.e. A scale which asks respondents to place a response to a question upon a predetermined range.

Figure 4–1 Example of a cognitive map based upon an interview with a state agronomist.

42 EXPLORING ENVIRONMENTAL CHANGE USING AN INTEGRATIVE METHOD


43

Behaviour occurs as a response to an interaction of attitude, situation and information that is continually updated and retrieved according to each new situation. Attitude, towards an object or action is therefore, only one determinant of behaviour (Fishbein, 1967) and the ability to measure the strength of an attitude does not necessarily correlate with any particular behaviour. Indeed the choice of behaviour to which the question(s) are directed may not even coincide with those that are considered salient by the respondent. For example work in the UK on the attitude of farmers to conservation has found that sympathy exists for the concept albeit with somewhat varied definitions about what constitutes conservation (Carr and Tait, 1991; Lemon and Park, 1993). However, when the concerns of those farmers are pursued it becomes apparent that the uncertainty engendered by agricultural policy has resulted in farming behaviour that has sought to maximise short term income with little thought being paid to conservation. This section has not been included to counsel against using attitudinal measures but to caution about reading too much into them, particularly if they are to be used in isolation without supporting evidence being supplied by other techniques. The strategy of ‘triangulation’ (Silverman, 1993) or multi-method approach (Perkins, 1988) which links different qualitative techniques recognises the need to estimate the thresholds at which particular behaviours may occur. However, it also highlights the need to observe behaviour and to extend George Kelly’s maxim beyond its’ psychotherapeutic roots “if you want to find out what is wrong in a client’s life, ask him—he may tell you” (Adams-Weber, 1981). The ability to observe and interact are fundamental skills of ethnography and we will briefly explore their value for issue based research. ETHNOGRAPHY AND CULTURAL INSIGHTS The insights from ethnographic studies about the cultural context within which confusing and often apparently contradictory behaviours may take place can be invaluable. For example, the status which is attached to farming activity can vary considerably (De Waal, 1991) and this will affect the willingness of family members to undertake farm work and can thereby limit the cropping options where a high labour component is required. Lemon and Green (1996) provide a further example of how ethnography can contribute to issue based research in a comparative study of how degradation processes are perceived in the Argolid Valley and the mountainous Epirus region of north western Greece. They argue that local rural understanding of “environmental degradation” is not defined simply by physical change but is closely linked to socio-economic processes. When changes in the physical environment were regarded as normal, or not connected to socio-economic life, they were not perceived as degradation. This was apparent in Epirus. In contrast, in the Argolid where physical changes were felt to be caused by socio-economic factors they were invariably considered as degradation. Where local socio-economic change was desired (Argolid), the “degradation” was regarded as inevitable and/or the responsibility of extra-local authorities rather than individuals or local communities. In the case of undesired socio-economic change (Epirus), “degradation” was regarded as having been caused by extra-local powers and events whose interests marginalised those of the local area. In both cases, degradation was locally constructed as being the result and site of struggles

44


between local and extra-local interest groups, where power was unequally weighted in favour of extra-local and particularly urban, interests. Such studies add a social and cultural dimension to the purely physical research into degradation processes. They have also established that processes which count as degradation are more closely linked to socio-economic and historical factors than to the physical processes themselves. Ethnographic insights can guard against the inappropriate analysis of data from more directed social enquiry and natural science-equally they can provide a cultural context definition and placement of a specific issue. This will be developed further in chapter six which draws upon semistructured interviews to establish a locally derived overview of natural degradation in the Argolid Valley. This overview is subsequently represented through a number of dimensions that determine the decision space(s) within which farmers operate. SOCIAL ENQUIRY AND THE MEASUREMENT OF NATURAL PHENOMENA The value of data about natural phenomena which has been acquired through social enquiry is often neglected by the natural sciences and undervalued by the social sciences. Such data invariably makes a number of assumptions that may not be readily acceptable to proponents of traditional scientific enquiry (i.e. the use of approximate data). It is, however, seen as supplementary and not in conflict with such a paradigm. There is a clear distinction between data that provides some measurement of the extent to which natural phenomena have changed and the process of scientific enquiry which is intent upon establishing the nature of those phenomena and the relationship between them. For example, information about the amount of salt in ground water in different places and at different times does not inform about the impact of salt upon various soil types and cropping regimes. Having made this distinction it is necessary to focus upon the characteristics of monitoring or the measurement of natural phenomena. ‘Good science’ would assume that this is accurate to the nth degree and is undertaken according to standard scientific procedures often requiring the employment of considerable technical expertise and accompanying technology. This traditional scientific model should be capable of providing an accurate picture about the state of a particular phenomenon at any one time, however it has a number of limitations that restrict its value for informing policy. Firstly, technical and personnel constraints limit the number of measurements that can be undertaken, at any one time or over time. Secondly this restricts the ability to obtain information about spatial variation. These limitations of scale mean that it is often impractical to provide data about the variation within an area and between areas. Finally once a natural process has been identified as significant it is often impossible to obtain historical data about it. Therefore, a complementarity can be identified between the acquisition of relatively few accurate (assuming scientific proficiency) measurements which cannot readily inform Table 4–1 Approaches and levels of definition.


45

about how that transformation contributes to the wider picture and some other way of collecting and presenting data which compromises accuracy but can inform in a more responsive manner and over a broader canvas (see Conway, 1985). A basic assumption underpins the use of social enquiry techniques, this is that the key actors in any system not only respond to the system as they see it but that they have a reasonable understanding of the degree of change in the natural world in which they operate. For example, a farmer will be able to inform fairly accurately about the amount of water he or she uses, the changing flows and salt content of that water and the depth at which it is accessed. The use of data from a sample of such farmers would enable the development of a picture of qualitative and quantitative change over a wider range of temporal and spatial scales than would be possible from the technical measurement of individual bore-holes over time. It is of course essential that some technical measurement is undertaken to validate, or otherwise, the data obtained from respondents. What emerges, therefore, is a data base that has different levels of definition appropriate to a range of functions (Table 4–1). METHODOLOGICAL PERSPECTIVES AND LEVELS OF SOCIAL ENQUIRY It has been seen that qualitative research: • • • • •

is based upon the need to understand the subject’s perspective places actions and meanings in their social context emphasises time, space and process uses everyday contexts rather than experimental conditions may adopt a range of data collection techniques.

Blaikie (1993) identifies a number of purposes for social enquiry: exploration, description, understanding, explanation, change and evaluation. These are obviously not mutually exclusive although they do demand a range of techniques and even disciplinary emphases. He argues that at the most fundamental level, social enquiry is; concerned with exploring some basic phenomenon that is not well understood, possibly to inform further stages of an investigation (p. 203). This exploration can progress little further than a description of the phenomenon and the circumstances in which it occurs or it can form the basis for an explanatory process which supports theoretical development and/or intervention at varying levels of intrusiveness. Either way the descriptive and exploratory phases are crucial (Murdoch, 1995) and should not be underestimated in the rush towards explanation and intervention. It is also important to recognise that any form of social enquiry, however sensitively it is handled, implies some form of impact or intervention in the process under investigation. Indeed it is the central feature of much of the participant observation and participatory research which is often adopted within community development work (Mikkelson, 1995; Okali et al, 1994). Any outside intervention will inevitably affect some local response and this carries with it certain ethical implications as well as practical ones. It is obviously highly questionable practice to carry out a study which implies subsequent action when this is not likely to be forthcoming. For example, an enquiry about the problems of water degradation and possible forms of remediation must be clearly

46


distanced from the ability to influence the changes necessary for that remedial activity. In the case of issue based research the primary objective is to establish how a particular issue is perceived as a formative phase in the identification of a range of possible futures. The need to establish the ‘decision space’ within which people operate is central to this strategic process and as such ‘empowers’ to the extent of information provision. While such an approach or method should be seen in the context of local participatory development with its accompanying goals of social justice, equity and democracy (Mikkelson, 1995) it is focused upon the need to provide a range of possible responses which incorporate physical and technical factors as well as those which are socio-cultural. In other words the output from issue driven research may well contribute to the options that are subsequently to be considered at a local level. FACTORS INFLUENCING CHOICE OF QUALITATIVE TECHNIQUE A number of factors influence the ability to access the study environment and to meet the enquiry objectives with minimal impact on that environment. There are two forms of access that need to be considered in the selection and subsequent implementation of social enquiry techniques, can you get there and will they communicate with you when you arrive? Cassell (1988) describes these as getting in (physical access) and getting on (social access). Physical Access Physical access is not just a question of reaching the study location but can also involve negotiating entry. It may be axiomatic to state that the ability to reach a study location is of particular relevance when that location is isolated. In such circumstances this will invariably coincide with the need to negotiate entry, possibly through a ‘gatekeeper’. For example, the interviewers had existing contacts in the central and western villages of the Argolid study area. These were interviewed where appropriate and were used to introduce other respondents when necessary i.e. where specific characteristics, such as farmers under thirty, were under represented within the sample. In many of the eastern villages the interviewers did not have contacts and as such were not trusted in the same way. This was overcome by drawing upon contacts who were also accepted in the other villages and were able to act as intermediaries. Similarly, entry into institutions is invariably eased, and occasionally only made possible, through the use of ‘gatekeepers’ who can ameliorate any mistrust. For example, entry into the local Agricultural Service was eased by the interviewers having worked as agronomists in the same office, however, the presence of a researcher from the United Kingdom and his lap-top computer immediately raised concerns about investigations from the European Union. In Greek public administration such mistrust can also manifest itself if the political affiliations of the researchers are, or are perceived to be, contrary to the study population (De Waal, 1991). The public administration is one example of a ‘closed access’ group which incorporate rules, procedures and individual gatekeepers as a way of ensuring secrecy for a variety of reasons i.e. client confidentiality, job protection (Hornsby-Smith, 1993). Professional associations provide another example of a closed group as indeed may be a village that can only be usefully accessed with the agreement of with the village head. Farmers in other villages however may be less influenced by its head and offer more ‘open access’ to the researcher.


47

A number of other issues are pertinent to the physical access of the researcher and should be taken into account. One of the major complaints levelled at the state agronomists by farmers in the Argolid was that they were only available in their offices during office hours. This meant that they were accessible at the very time when the farmers were committed to working on their land. Similarly, it was inappropriate to attempt to interview farmers at the time when they were at work and particularly when the work load is very heavy i.e. during the harvest. Research in the United Kingdom and the United States (Buttel, 1989; Lemon and Park, 1993) has suggested that farmers are willing to be interviewed in their home environment. This was often not the case in the Argolid study and the interviews invariably took place in the village cafes where the farmers congregated to talk, drink and play cards. While the cafes provided a good location for meeting groups of farmers and for ‘snowballing’ contacts i.e. using one interviewee to introduce another, they also highlighted a number of sampling considerations. Two examples of which can be provided by the Argolid study. Firstly, the cafes tended to be dominated by middle aged and older men and as such the younger male farmers and particularly the female farmers were not available. Interviews with these groups were carried out in other locations, usually the farmhouse. Secondly, political allegiances are often village based and sometimes vary from cafe to cafe within one village. This example highlights the importance which is attached to politics within the culture and thereby its ability to colour the responses to an interview particularly when it is held in the presence of others. At a more general level it represents the need to be conversant with the cultural context within which the research is to be undertaken. While this should be part of the ethnographers ‘tool box’ it is of equal importance when undertaking more directed ‘issue based’ social enquiry. Social Access Physical access to a research group is inseparable from the ability to negotiate social access and a recognition of the areas in which such access might be problematic. There are circumstances, however, in which the process by which physical access is obtained can result in social access or acceptability being subsequently questioned or with-held. For example using the Director to sanction research into the Service of Agriculture without negotiating entry at the level at which the work is to be carried out. The politics of mistrust (Lee, 1992) can provide a considerable constraint to social enquiry particularly when it is insensitively handled. However, it can be apparent even when the researcher has been empathetic and sensitive and has negotiated entry to the research environment but is experiencing a negative or ambivalent response. In such circumstances the existence of a ‘sponsor’, or intermediary, who can reduce distrust and help the researcher to gain acceptance over a period of time, is invaluable. A classic example of this is ‘Doc’ in Whyte’s study of a street corner gang (Whyte, 1955). In the Argolid study this sponsor was invariably the person who had provided the initial contacts within the village (i.e. the village secretary), in some cases however, they emerged after that initial entry had been made. The fieldworkers for the research in the Argolid were accepted in many villages because they had previously established contacts and trust through their work as agronomists with the Service of Agriculture. Their most recent project, prior to the current research, had been to oversee the subsidised programme for uprooting the Apricot trees which had been decimated by the Sharka virus. As has been seen this role could have alienated the researcher from the farmer if the relationship had been ‘formal’, office based and put the emphasis upon the farmer to acquire the services of the

48


agronomists. Both investigators had deliberately made their work field based and not contingent upon office hours. While this caused some problems within the Service which required staff to clock on and off it engendered trust within the farming population. Similarly the researchers were not felt to be a ‘threat’ either from central government in Athens or from the European Union in Brussels. In order to benefit from the relationship between the fieldworkers and the farmers some sacrifices had to be made concerning the ‘disciplinary’ orientation of the research. In other words no expectation was made of the researchers to act as sociologists or anthropologists. While this was felt to be appropriate for the more directed form of social enquiry required it also meant that most of the ethnographic insights had to be acquired independent of the field work activity. A number of other features are relevant to social access, some of these may appear superficial but their impact should not be underestimated. It has been seen that the location for the interviews carries a number of cultural implications i.e. arising out of political allegiances. They also imply a social etiquette preceding the conduct of an interview. This invariably took the form of protracted conversations about politics and football over food, coffee and copious amounts of beer and ouzo until the early hours of the morning. It was only then that the farmers were willing to proceed with the interview. Therefore, social access was dependent not only upon knowledge about the substantive issue—agronomy and degradation, but about the culture and interests of the respondents, and about the potential for conflict. In other field work situations i.e. interviewing scientists or agronomists, social access may be contingent upon such factors as the need for confidentiality, comprehension of technical language and the culture and procedures of the profession. This is not only relevant to accessing the fieldwork environment it is also pertinent to the knowledge that is required to undertake that activity. Knowledge Two forms of knowledge are central to the choice and implementation of social enquiry techniques. Firstly, it is necessary to have a clear understanding of the potential and constraints of the techniques available and the skills that are necessary for their use. Secondly, and following on from the previous section, sufficient knowledge must be available to the researcher about the field work environment to support physical and social access and the subsequent undertaking of research. There is an extensive social science literature on the subject (i.e. Blaikie, 1993; Silverman, 1993; Foddy, 1995; Gilbert, 1993; Yin, 1985) and this is supplemented by texts that report the research process, warts and all, and as such provide an invaluable insight into fieldwork procedure and the techniques adopted (Bell and Newby, 1977). Knowledge about the research environment has already been considered as fundamental to access, it is also of crucial importance to the way that the research is undertaken and, of course, interpreted. The language employed in more structured forms of enquiry and the format of that enquiry is culturally specific. For example, an attempt to elicit how farmers perceived risk in the Argolid was firmly rejected by the field workers because they felt that many of the interviewees would want to be discursive in their responses and others would be alienated by any ‘quantification’ of those responses. Therefore, what was felt to be an ingenious set of questions, designed to elicit the ‘probability’ of different futures, had to be rejected because they were deemed to be inappropriate to the research environment. Pilot studies are also useful for exploring whether the language and techniques are understandable to the respondents, however with small samples some care has to be taken not to undermine the goodwill of the respondents.


49

While care must be taken to use language and terminology which is understandable to the respondents it is also important to recognise that their own language may not coincide with, or be comprehensible to, the researcher. Research by Lemon (1991) in an English village found that some contacts used terminology in earlier meetings that was drawn from local dialect but seldom used in contemporary conversation. Two explanations can be put forward to explain this and are evident, in various forms, in many other studies. Firstly, the respondent may be acting out the role which they feel is of interest to the researcher. Alternatively, they might be testing the researcher to see whether they are aware of the terminology and if not whether they are prepared to admit their ignorance and enquire about its meaning. One farmer in the Argolid had entered two elephants in his livestock returns for over a decade. This was less of a deliberate attempt to mislead than a statement about the intrusion of the state in farming activity and as such says something about the desire for independence within the farming community. The use of a sponsor and or locally accepted researchers can reduce some of these communication problems. Similarly with the undertaking of more directed ‘issue based’ research the more subtle linguistic and cultural nuances may be overlooked. It is important therefore to establish and maintain these contacts and to be aware of the insights that can be gained from ethnographic studies. SOCIAL ENQUIRY RESEARCH TOOLS—INTERVIEWS AND QUESTIONNAIRES Much of the data obtained for this study was collected using social enquiry research techniques, primarily semi-structured and structured interviews. The contribution that such techniques can make towards an improved understanding of natural phenomena and human-natural interactions have been presented above as have some of the factors which influence the choice and implementation of these techniques. This section will discuss some of the general characteristics and procedures relating to these techniques and will draw upon the Argolid case study to exemplify these. Chapter two highlighted three interfaces for linking ecological and agricultural systems (agroecosytem, farmer decision making and policy making). It was also argued that an appreciation of the range of stakeholders and the variety of perspectives on the system under investigation is essential for developing an holistic view of that system. In terms of data requirements and the techniques available for collecting this data a number of factors need to be addressed. Firstly, exisiting data sets need to be identified and accessed i.e. census and meteorological data. Secondly, where the information aquired from those data sources is not at sufficient resolution other approaches have to be considered. For example do agencies collect data at a more local level or is it necessary to undertake some form of primary data collection exercise to obtain it? Thirdly it is important to be able to distinguish between the individual and organisational perspective (Linstone, 1981). It is not possible to interview the local Service of Agriculture although the Director may be accessible. Similarly it is not possible to interview the Department of Agriculture or an EU Directorate although representatives may be available. Even where representatives can be reached they can only provide an individual interpretation of the organisational position. In consequence it is often necessary to draw upon official documentation either to compare with the respondents’ interpretation or as the only insight into an organisational perspective. Table 4–2 outlines the range of data that was collected for the Argolid study and the source of that information.

50


As can be seen from Table 4–2 much of the data for the study was collected from interviews with key actors in the area. These were of two types, firstly, an initial set of semi-structured interviews were carried out with farmers, agronomists, politicians, public servants, co-operative managers and scientists to establish how the agricultural production system was perceived and to what extent the condition of the water system affected this perception. This formed the basis for understanding the different agendas to which individuals and groups were operating and the perspectives that were implicit within these. For example the ‘technical perspective’ adopted by the scientific community is discussed in chapters nine and ten. These interviews also established the relevant attributes and parameters for a subsequent, more structured, set of 203 interviews with farmers distributed throughout the area. These were designed to establish the extent of water degradation and the factors and thresholds that determine crop choice i.e. crop price, water costs, existing farming activities and structures. Information was also collected concerning the perceived risk and uncertainty in agricultural production and the costs and income related to that production. Table 4–2 Data requirements and source.

The spatial distribution of crops, water availability and water quality was investigated alongside the variation in income that results from farming different locations with a corresponding range of soil types, topography, micro-climate etc. These attributes were classified spatially through a seven zone format, the basis for which is discussed at the end of this chapter. However, before this the


51

decisions taken, and procedures followed, for the Argolid field work will be considered and used to exemplify more general points about this type of social enquiry. STUDY INTERVIEWS Interviewing involves the direct questioning of individuals or groups about a specific phenomenon or phenomena. The structure of an interview will vary according to the pre-determined agenda of the interviewer and will range from the use of highly structured questions to open ended and loosely structured ones. Before an interview based study is undertaken a number of points need to be borne in mind. 1. What information is being sought from the study? 2. What are the sample populations and how are they likely to respond to different approaches? 3. What form of interview technique is appropriate? 4. Who will carry out the interviews? 5. What level of interviewer training will be necessary? 6. How many interviews are required and what form of sampling will be undertaken? 7. What will be the cost of the field work and equipment required? 8. How will data processing and analysis be undertaken? The information sought from the study has already been differentiated between that which supports a broad overview and that which provides more structured data about agricultural production and water use. Similarly the sample population has been identified as a range of key actors for the exploratory phase of the work and a spatially representative sample of farmers across the Argolid for the second phase. In both cases the interviews were undertaken by local agronomists who were knowledgeable about the area and agricultural issues and were sympathetically received by the farming community. Although they had previously worked for the local Service of Agriculture and had established a wide range of contacts in so doing, the interviewers were not seen to represent either a political party or a sector of the public administration. This meant that many of the concerns discussed in the previous sections about social and physical access could be addressed. The selection of local ‘experts’ enabled much of the design phase to be piloted within the study team. For example the relevance of questions and the way in which they were constructed was often challenged by the field workers. Similarly, attempts at relatively sophisticated risk assessment exercises were dismissed out of hand because they were felt to be alien to the way of thinking within the farming community, the time needed to undertake the tests was too great and the type, and level, of respondent numeracy i.e. for estimating probability was insufficient. One of the interviewers spoke excellent English and was therefore able to translate and transcribe the responses3 although in the first phase of the study this was carried out alongside the second interviewer, who was more experienced in local agricultural issues, and the team’s social scientist who ensured that an accurate overview was established. This interactive process also limited the distortion that could occur by moving from Greek to English and from the ‘over interpretation’ of responses by the interviewers. Although on balance the use of known local experts was invaluable, the danger of influencing respondents and interpreting output had to be guarded against. A more pragmatic

Figure 4–2 Schematic example for exploring the issue of depleted water stocks in the periphery.

52 EXPLORING ENVIRONMENTAL CHANGE USING AN INTEGRATIVE METHOD


53

assessment is that without the contribution of this type of interviewer problems of ‘access’ and trust would have severely handicapped the second phase of the work, interviewing sceptical farmers. The training of interviewers was therefore directed at how to pursue a line of questioning or to instigate a new line without influencing, too greatly, the content of the response. This was obviously a more significant requirement in the exploratory phase of the work and a simple guide was to explore issues as processes. This meant that when an issue was raised its’ perceived history could be explored through establishing perceived causal linkages. Similarly, the anticipated impacts resulting from these processes could be elicited without substantively directing the nature of the response. A more formal description of this pathways approach is presented in Lemon (1991) and Lemon and Jeffrey (1998) and a schematic representation is shown in Figure 4–2. By exploring issues in this way the respondent not only highlights process interactions but also agencies (social, economic, political etc.) and the level at which they operate. Preliminary discussion among the research team, including the interviewers, highlighted a number of broad issues that had emerged from the exploratory interviews and could be pursued further. These focused upon the changing character of agricultural production, the perception of degradation and attitudes towards remediation and the identification of agencies and temporal/spatial scale over which these processes had occurred. Concepts such as scale were not intended to be elicited directly but implied from the processes as they were interpreted. For example in Figure 4–2 the period over which rainfall was reduced could be established and verified and the level of irrigation water compared between the periphery and the central plain. The latter of these required additional more structured data which was collected in the second phase of field work thereby highlighting the benefit of a multi-technique approach. The characteristics of these approaches will now be considered in more detail. A fundamental difference concerns the point at which ‘bias’ is introduced into the procedure. In the structured interview this will be at the design stage whereas in the unstructured approach it will occur primarily at the point of analysis. A practical compromise (and one that is frequently adopted) is the semi-structured interview which combines the flexibility of more open questions with closed questions which can often provide a quantitative data set. Unstructured Interviews As a general rule the less structure that is imposed upon an interview the more potential it has as an exploratory tool (Table 4–3). This format requires considerable interviewer skills i.e. an ability to combine sensitivity and empathy with a clear understanding of the objectives of the study. These interviews are closer to natural conversation, however in order to meet the objectives of a study they must focus upon particular subjects, processes or issues even if these are relatively abstract or obscure. Figure 4–2 provides an example of how the results of such an unstructured interview can be presented in a structured format. In phase one of the Argolid study the interview format was built around a check list of issues to be explored—i.e. how has agriculture changed in the area over the last forty years? Supplementary questions would be asked in order to further the exploration. This obviously places a considerable responsibility on the shoulders of the interviewer, however, it also allows for clarification on a point when necessary. From the point of view of the interviewee 3 The unstructered interviews carried out in phase one were tape recorded although hand written notes were also made. The time required for the transcription process and analysis should not be underestimated and between eight and ten hours for an hour length transcript would not be excessive (Bertraux, 1981).

54


Table 4–3 Advantages and disadvantages of less structured approach.

this approach allows them to respond in their own terms and this again requires the interviewer, and subsequently the analyst, to be aware of the meaning attached to language. In the case of the Argolid study, however, no attempt was made to undertake linguistic analysis beyond interpreting the ‘essence’ of what was being said. A literal translation was sought, therefore, rather than one which exposed cultural nuances or hidden agendas. A less structured approach to interviewing will tend to employ open ended questions that allow considerable freedom in response. By contrast a more structured approach will focus more on closed questions which direct the response into a pre-ordained classification. Structured Interviews Table 4–4 outlines the advantages and disadvantages of structured interviews and questionnaires which can be administered or self-completed.4 These focus on the removal of spontaneity in return for the opportunity to replicate and compare. The link between questionnaires and interviews in this context is that a more structured interview is in effect an administered questionnaire—similar to those employed by market research firms in shopping centres etc. In the context of the second phase of the Argolid study the interviews were administered by local agronomists and were divided into four sections consisting of introductory or background questions (size of farm etc.); climate and natural hazards affecting the way in which the respondents farm; cost and income from farming and water use, quality and costs. A detailed representation of the questions asked in the structured second phase of the field work is provided below. While social enquiry in general has been seen to often consist of multiple techniques used in a co-ordinated manner it is unlikely that any individual tool will be either totally structured and closed or unstructured and open. Even when no other restrictive responses are sought then some form of comparative information will generally be required about the respondents. This allows for comparison within the sample, between that sample and like studies and for establishing how representative that sample is of the total population. This latter point is important because it allows one to adjust a sample so that it is representative or if this is unnecessary or

4

These are often delivered by post.


55

Table 4–4 Advantages and disadvantages of less structured format.

impractable to interpret the findings in the context of the sample used. By extension therefore unstructured and open enquiry techniques tend to be restricted to certain types of ethnographic investigation whereas a more structured, albeit not totally structured, approach might be more appropriate elsewhere. In effect therefore there exists a continuum from the totally structured format to that which is completely open with the majority of interview designs being semi-structured and falling somewhere between the two with their respective advantages and disadvantages. Group Interviews With some of the exploratory interviews and the more discursive elements of the more structured second phase interviews it was, on occasions, propitious to interview more than one respondent at a time. The reasoning behind this approach in the Argolid was that response was more likely when a group were together i.e. in the village cafe, than by trying to separate the group. More widely however, group interviewing has become standard practice in market research whereby carefully selected samples are brought together to discuss a product or the market place for a product. Similar techniques have also become a widely used weapon in the armoury of political parties as they strive to establish the issues that are important to voters and how those voters might respond to particular policies. In the Argolid study, however, group interviews tended to equate to sampling serendipity

56


rather than elicitation technique. In either case the approach requires considerable skill on the part of the interviewer(s) and it is advantageous to have two interviewers, one of whom is responsible for the shape of the interview and one to record both the verbal and, where appropriate, non-verbal responses. Table 4–5 highlights the advantages and disadvantages of group interviews. Telephone Interviews (Fax, e.mail etc.) Other interviewing media are becoming increasingly popular, particularly for organisation and work based interviews where there is usually a phone (or fax for questionnaires) and the potential Table 4–5 Advantages and disadvantages of group interviews.

Table 4–6 Advantages and disadvantages of using telephones, faxes, E.mail etc. for interviews.

respondents are often too busy to commit themselves to a meeting. The use of the telephone can also support an initial enquiry phase intended to determine the focus for subsequent phases and/or to help define a sample for face to face interviews. In the Argolid study telephone contact was used to arrange some of the exploratory interviews although relatively little information was obtained in this way. Ongoing contact has however, been retained with some of the original respondents (i.e. the Young Farmers Association), initially by fax and more recently by E.mail. While this has been useful for updating information it reinforces one of the problems attached to such media for wider interviewing purposes in that it only provides access to a restricted sample with a technological orientation that is not typical among the farming population as a whole. Indeed it is this orientation that is seen by other farmers as one of the few ways that they can access information.


57

Question Design and Piloting The questions asked in the second phase of the study are discussed below. Prior to this, however, it will be useful to introduce some of the basic design caveats of questionnaires and structured interviews. A considerably more comprehensive coverage is provided by Foddy (1995) who identifies a number of pitfalls in questionnaire design and the phrasing of questions (pp. 2–9). • Factual questions can elicit invalid answers: This may be due to an oversight on the part of the respondent, alternatively it could be due to the respondents unwillingness to impart specific information (i.e. exaggerating the amount of land used to grow a subsidised crop or underestimating the water used for irrigation when restrictions are in place). • The relationship between what respondents say they do and what they do is not always very strong. This can be extended to question the relationship between attitudes and behaviour (Miller, 1980) both of which can be highly unstable. For example, the data relating to the anticipated responses to different types of perceived risk (i.e. drop in price support, an extended period of drought or heavy frosts) which is discussed in Chapter nine, must be handled carefully. • Respondents often misinterpret questions or interpret them in a way that is contextually or culturally specific. This has already been considered in the earlier section on ‘social access’ and is a two way consideration. The culture of the researcher as expert and the use of technical language can lead to misinterpretation or discomfort in much the same way that the use of particular terms or views by the respondent may be ambiguous or misleading. • Changes in the order of questions can affect responses. If the respondent is directed towards the particular phenomenon of interest to the researcher then it is likely that the wider context within which that phenomenon is situated will be lost. For example, to focus immediately upon the quality of irrigation water will allow for detailed questions about that issue. It is less likely, however, to be of value in contextualising that issue within the agricultural production system as a whole. In other words if the initial questions relate to water quality then it will be difficult to establish how important that issue is in a wider context. It is possible, therefore, that the issue will be of little overall importance (compared to prices, climatic conditions etc.) but detailed responses may be obtained that reinforce the researchers view that it is a central issue for farmers. • Different question formats (i.e. open and closed questions) can affect the response. This has already been discussed in terms of the ability to explore a response in a less directed way or to fit that response into a predetermined format developed by the researcher. To develop the above example, an open question about what a respondent considers to be the most important issues facing local agriculture may elicit a number of responses but not that of poor quality water. Asking about water quality and questioning whether or not it is an issue will undoubtedly raise the profile of that issue and influence the response accordingly. • Respondents can supply answers when they know little about the topic. This is often less of a deliberate attempt to mislead than a positive attempt to contribute. These points reinforce Moser and Kalton’s (1971) emphasis upon the need for a questionnaire or interview to be relevant to the respondent. This is implicit in the discussion on access and can be modified by saying that social interaction with a respondent may increase the relevance of the study for them through sympathy for the interviewer as much as the subject matter. While this is invariably a good thing because it helps to ease the passage of an interview it can also be dangerous

58


in that the desire on the part of the respondent to prolong the interaction may over-ride that for supplying ‘accurate’ responses and as such necessitates that the interviewer distinguishes between the social and the professional components of the interaction. A number of other considerations about the design and phrasing of questions relate to the need for clarity. Although this is of crucial importance in self completed questionnaires it is also a key consideration for more structured interview formats and again highlights the need for knowledge about the sample population. For example, care must be taken to avoid technical language that will not be understood by all the respondents. Similarly consideration must be given to terms that may have multiple meanings i.e. within a cultural context or more simply through ambiguous wording. For example the terms ‘building’ and ‘development’ can be seen to have specific meanings if the latter is defined more clearly as economic or community development. When this clarification is not provided the term development can easily be interpreted as the provision of new housing or industrial units and as such may be confused with building. In short the clarity of an interview or questionnaire design is dependent upon having the same frame of reference as those under study (de Vaus, 1996). One way of ensuring this is to ‘pilot’ the design. PILOTING Piloting or pre-testing is particularly important with structured data collection because once the study is under way it is more difficult to adapt than a less structured format. The piloting exercise will normally be undertaken with a sample similar to that intended for the main study and attention will be paid to any ambiguities or lack of clarity in the design and language used, or in the type of responses. In the second more structured phase of the Argolid study piloting took two forms. The interviews were tested in the normal way with twenty farmers and small adjustments made as a result, for example it was decided to concentrate upon details about the main crop grown by the farmer rather than pursuing each crop individually. While this restricted the information that was made available it shortened the interview length while focusing upon the water, farmer and crop characteristics for the majority of the farmed land. However, as can be seen below from the questions asked, basic information was obtained about the location of all crops. Prior to moving into the field a pre-pilot was carried out in which the questions were scrutinised by the interviewers who were knowledgeable about the farming population. This led to the removal of a section of questions designed to elicit perceived probabilities to evaluate the likelihood and magnitude of a range of risks. It was argued that the devised technique involved a form of numeracy (i.e. the use of probability) that would be alien to many of the farmers and even if it was not would distract from meeting other, more important objectives of the interview. This pre-pilot also highlighted some of the problems of access and in so doing helped to clarify some of the sampling issues which will now be discussed. SAMPLING Sampling is an ‘assurance’ that the respondents are representative of the group under study when it is impossible to cover the total population of that group. The achievement of such a ‘representative’ sample can however be misleading. For example the random selection of possible respondents can become meaningless if those that do reply do so because they have a vested interest in the issue concerned. It is more realistic to be aware how representative a sample is, and if necessary, to adjust


59

it to include those groups that appear under-represented. For example, as has already been discussed, the interviews undertaken in the second phase of the study were carried out in village cafes and under represented both young and women farmers. It was therefore necessary to arrange meetings with women farmers at other locations, generally in their homes. The representativeness of a sample can be partly established using secondary data for comparative purposes. For example the age and gender distribution of farmers in the Argolid was established from the census and compared with the distribution of respondents in the sample. A number of sampling approaches can be adopted, some of which are outlined below. However, a comprehensive literature exists on the subject and a good introduction is provided by Arber (1993). Probability Samples i. randomly generated: i.e. by post code, membership number, ii. systemic sampling: i.e. every twentieth member. iii. stratified sampling: i.e. a sample of males or one consisting of the same male: female ratio that exists in the total population. iv. cluster sampling: i.e. the random selection of a particular grouping and a random sample taken from within that selection. Non Probability Samples i. Judgement samples are based upon the researchers knowledge of the ‘world’ in which the respondents operate i.e. active in environmental groups. ii. Opportunistic sampling is based upon those who are willing to co-operate, iii. Snowball sampling is based upon contacts provided by existing respondents. The sampling undertaken for both phases of the Argolid field work fell into the non probability category with judgement samples and snowballing techniques used in the first phase by drawing upon the local knowledge of the interviewers and the contacts of the interviewees. No attempt was made to establish a representative sample in this exercise because of the range of possible interviewees (farmers, scientists, public administrators etc.) although representation was sought from each set of actors, or stakeholders, who were considered relevant to agricultural production and water use. In the second, more structured phase of the research similar techniques were adopted alongside an opportunistic component in which respondents were found in the cafes and other social settings. As a result the exact samples would be difficult to replicate without the involvement of the same interviewers. CODING AND ANALYSIS It is essential that some thought is given to the handling and managing of data at the design stage. The piloting process should have uncovered some of the discrepancies in the questionnaire or interview both in their implementation and in their subsequent coding. To be of value, over and above the purely descriptive, it is necessary for some form of comparative criteria to be imposed upon data. As has already been seen above this will occur in the design of the research instrument

60


for more structured formats i.e. where the range of responses is given. Where there is a more open ended approach this judgmental stage will occur after the data has been collected, however, the clear setting of research objectives will often impose some form of comparative structure. We will only consider coding procedures in as far as they have been adopted within the Argolid study. This provides examples of handling data which have been collected from a number of enquiry techniques (i.e. text analysis, structured and semi-structured interviews). It will be useful, however, to introduce a number of general rules for the coding process (Fielding, 1993). Firstly codes must be mutually exclusive, (i.e. while a farmer may grow olives and oranges at the same time, she cannot be both married and single). Secondly they must be exhaustive. All possible options must be covered although some aggregation will inevitably occur, after all this is the basis of the coding process (i.e. oranges, lemons and grapefruit may be coded as citrus and other crops as other). Ideally the piloting process should establish the range of possible responses. It is extremely irritating when new phenomena emerge that have to be fitted into an existing coding structure for which they are not suited. Alternatively that structure has to be adapted in such a way that it becomes harder to manage or the miscellaneous category grows to be disproportionate in size and thereby meaningless. Finally any coding structure must be applied consistently throughout and with qualitative data this may well require some form of arbitration or second opinion (Knippendorf, 1980). The coding process has a number of stages which we will examine briefly in the context of the data collected for the Argolid study. There is obviously a difference in the approach adopted for questions which have a ‘closed’ format and as such can readily be attributed coded values from that undertaken for open ended questions which require more interpretation. The use of a standard statistical computer package such as SPSS (Statistical Package for the Social Sciences) can be employed to enter this data and to run the standard descriptive statistics that may be required. It is also capable of more sophisticated tests but with the data collected for the Argolid project these were considered inappropriate and caution should always be exercised to avoid ‘throwing’ statistics at data. More structured questions can elicit information at intervals along a continuous scale (i.e. when were oranges first grown on your farm?). Alternatively an ordinal scale can be used for responses which select from a predetermined set (i.e. in 1994 do you expect the amount you spend on water to 1: increase a lot ..to..5: decrease a lot) or on a nominal scale for which there is no intrinsic ordering (i.e. do you sell your crop to the co-op, to the local market or is it for domestic consumption?). In the context of the coding sheet these could be represented as follows. Table 4–7 Coding sheet for interval, ordinal and nominal data.

Even with pre-coded questions it is often useful to retain the option for responses that have not been anticipated (i.e. “other” under market in Table 4–7) however the need for this should have been minimised through the establishment of a clear understanding about the system under study


61

prior to data entry. This uncertainty is far more difficult to pre-empt with open ended responses and text. For the Argolid study the qualitative data obtained from the first research phase provided the basis for the more structured questions in phase two. Each interview was broken down into discrete statements, each one representing one concept (an activity, an attitude, an innovation etc.). For example a brief extract from an interview with the wife of a farmer who, unusually for this area, has adopted ecological farming techniques on one of his farms. The neighbours react to Spyros’s ecological approach with indifference They panic when they see a disease They rely on the agronomists who sell the pesticides The agronomists have a negative view on ecological farming Because they want to sell These statements were then coded according to their descriptive classification or topic (i.e. water quality, public administration, attitude to farming, transfer of information) and where necessary more than one classification applied to one statement. For example the statement ‘they rely on the agronomists who sell the pesticides’ tells us something about how this respondent perceives information flows, attitudes to farming and response to uncertainty and the procedure for purchasing inputs. It is the intention of coding to link units of data and it is the nature of these linkages that can help move towards a systemic picture of an issue (Figure 4–3). Figure 4–3 shows a number of linkages from the perspective of one respondent around the concept of ecological farming. This supports the possibility of a substantive and comparative interpretation around that concept. In other words what are the related issues perceived by the respondent to organic farming and how are these issues seen to link. For example the concept of ‘traditional farming’ which is accredited to Spyros’s father is defined by high input techniques and the ‘new’ approach is much closer to what elsewhere might be considered ‘traditional’. If this map was extended it could be seen that one of the distinguishing features of this ‘modern’ group of young farmers is their age, educational background and use of information technology. The example uses statements taken directly from an interview and as such is one representation of rich data. It is a time consuming approach and not always practical, particularly with large numbers of respondents and or responses, to present data in such a way. The exploratory process whereby the analyst gets ‘a feel’ for the data and the rich picture associated with it, is however, extremely important and should preceed any attempts to classify and structure qualitative data into comparative formats. The coding and subsequent analysis of open ended and textual data can take two broad forms which can also be stages in an iterative process. Firstly the data can be examined for common themes and concepts, this could be undertaken using Boolean search techniques or a simple word search. What is most important to note is that no computer package can undertake the analysis for you. The questions that are asked of the data and the interpretation of that data must be the result of becoming immersed in it. Reading and rereading the interviews, and where necessary clarifying points that are unclear, is an essential prerequisite to the coding and subsequent analysis of qualitative data. ‘Trawling’ the data with the latest computer software will only rarely throw up insights that the process of immersion has failed to identify. Once the issues and categories for analysis have been clarified and a coding system established i.e. searching for specific words or attaching numeric codes to concepts or themes then the use of computing power is invaluable, particularly with large data bases.


Figure 4–3 Example of qualitative linkages.

62


63

By establishing some form of descriptive ‘identifier’ against individual statements it is then possible to sort the text according to that theme and to establish an overall picture, across the sample, of statements which refer to it and to provide a comparative subset drawn from individual interviews. Obviously most statements can be classified according to more than one descriptive criterion. For example the temporal (duration and velocity) and spatial (geography and organisation) characteristics may be considered alongside some qualitative assessment i.e. whether the statement refers to a positive or negative concept. If this approach is extended then the pathways analogy shown in Figure 4–2 can be adapted to indicate an analytical framework which essentially moves from the abstract level (spatial and temporal scale, social networks and interactions), through a number of sub-classifications (farming practice, social structures and attitudes, sources of information etc.) to a descriptive and often idiosyncratic level similar to that represented in Figure 4–3. In reverse, as the descriptive data is classified in a more abstract manner it becomes increasingly possible to establish comparative linkages. A more detailed exposition of this approach to handling qualitative data is given in Winder and van der Leuuw (1997). We will now focus upon the structured interview which provided complementary information to that obtained from secondary sources (i.e. census figures) and was used as the basis for the decision making model (Chapter 10) and for establishing the representative zones adopted for the models. INTERVIEW STRUCTURE FOR PHASE TWO As we have already indicated the second phase of the Argolid field work was undertaken through a structured interview or administered questionnaire. The questions that were include in the interview were broken down into four main sectors. Introductory questions that provided background information about the respondent farmer and their agricultural activity; the uncertainty that was perceived in terms of climate and natural hazards; anticipated changes in price and the perceived responses to these and finally uncertainty relating to water costs and quality. Part 1: Introductory Questions Where do you live?: Since intervews took place through an informal networking process and mostly in public places late in the evening it was important to identify the home village of the respondents.5 How many parcels does your farm consist of?: The territory selected consisted of 39 villages with considerable variation in the number of farmers, the size, number and dispersal of plots they farm and what crops are grown in these locations. For each parcel of land the following information was sought: The location of the parce by village

5

Farmers were generally inaccessible during the day and interviews took place during their recreational time in the cool of the evenings after day time temperatures in excess of 35°C

64


Size of parcel Current crop grown on that parcel Where that crop was sold Overall, this data can provide the proportion of different crops grown in villages and zones and the markets to which they are sold. The predisposition of farmers from diferent locations to grow certain crops and to use particular markets was also sought. Apart from farming what other sources of income do you have? What other sources of income do other members of your household have? (Wife, Children, Others) The intention here was to provide a link to the sociological investigation which suggested that there are varying degrees of economic dependency on agricultural income and that this might have a spatial distribution which could provide some indication about the feasibility of different cropping options across the area. Which of the following jobs do you do for yourself on your farm? If you do not do them who does? Pruning; Fruit picking/harvesting; Fertiliser application; Irrigation; Pesticide applications: Book keeping (accounts) This question was intended to add further data to the typology of farmers and to differentiate crops by labour content and cost. It also identifies in book keeping an issue which is the subject of policy on managerialism in farming. What is your main crop? Do you co-cultivate this crop/with what? What is total production in kilos of your main crop? Where appropriate—What is the average production per tree? From here on the questions concerned only that crop which the farmer nominated as his main crop. It had been found that there were many instances of farmers not only working a number of parcels of land but growing a range of crops on different parcels. An attempt to design an interview which would give information on all of these options proved far too complicated to negotiate with farmers. The compromise was to concentrate on the crop which farmers nominated as most important to them and then to focus on those parcels of land on which that crop is grown. Some initial ambiguity occurred about what was perceived to be the most important crop, should it be defined by income or area, the former option was subsequently adopted. A knowledge of total production for each respondent, by weight of that crop, enabled total production by crop type to be estimated by village and zone. The variability in this production could then be related to the nature of the farming activity, input resources and location specific factors such soil, irrigability and cultivation environment. The question was also intended to provide a comparative data set to that which had been acquired from official sources. Last year did you use artificial rain (sprinklers), free flow or both for irrigation? How many times did you irrigate this crop last year? For how long did you irrigate each time? What is the rate of water flow for this irrigation? Did you use artificial rain against frost/How often?


65

The first part of this question differentiated between the more technically sophisticated method of irrigation using sprays and the more traditional method of periodic flooding. Irrigation specialists have a great deal to say about the relative merits of different techniques with respect to the efficient use of water (Van Tuijl, 1993). There is also a considerable difference in the cost of equipment. In other parts of the Mediterranean where water is explicitly priced more sophisticated equipment is used to administer doses of water directly to the root system of many crops. The next three questions enabled the amount of water used for irrigation to be estimated and combined for villages, crops, and zones. They also supported an independent comparison with field data derived from hydrological measurement. The final question was stimulated by the increased use of fine sprays during nights of potential air frost. From where did you obtain the water for this irrigation?: There was clearly a need to differentiate between the sources of water used for the main crop and the plots on which it is grown. The sources have different costs and variable quantity and quality and since water itself is not, in general, priced the cost to the farmers is one of access. If you have bore-holes, how many do you have? Do you own these yourself or with others? How deep are these holes? What is the flow from these holes? When (date) did you first use your own bore-holes? A large proportion of agricultural water has been obtained from wells and bore-holes. From the 1930’s until the 1960’s water in the central plain was primarily obtained from conventional wells. As the water tables in the aquifers dropped from the 1960’s onwards and drilling equipment became available in this part of Greece farmers began to drill deeper for water. The cost of water from boreholes is a function of the capital cost for the equipment (pump etc.) which in turn is determined by the depth of drilling and the operating costs (i.e. electricity) of the pumps. Additional costs were incurred when new bore-holes failed to yield water or when the failure of an existing bore-hole precipitated the need to drill another. This failure was due either to the drying up of the bore-hole or to the salination of its water. Information was obtained on bore-holes owned or used by respondents. This included when they were first drilled, their depth, the initial flow, and whether and when they had failed, and why. As a consequence it has also been possible to analyse this data and move towards the development of a strategic water use and water/salt model over the last fifty years (Chapter 11). This in turn can be compared with the hydrological data collected by the Agricultural University of Athens. Indeed, this part of the interview provided some data which it had not been possible to obtain through conventional hydrological measurement. Do you have the use of any of the following for producing this crop? Electric pumps Air mixers Water filters Bore-hole isolations Irrigation system Artificial rain

How many How many How many How many metres of pipe

Date first used Date first used Date first used Date first used Date first used Date first used

66


This data gives a further estimate of capital equipment used and provides a history of technology diffusion to be developed as part of the temporal analysis of water intensive cropping (Chapter 7). Part 2: Climate and Natural Hazards Which of the following do you think is the greatest threat to your main crop? Less rainfall; Frost; Pests/viruses; Other In an attempt to understand how farmers handle risk and uncertainty it was felt necessary to identify what factors they perceived as threats yet were outside of their control. Climatic conditions and the existence of crop disease and pests were the two most obvious factors upon which to focus. In 1994 do you think this will be: Much better, Better; No change; Worse; Much worse Ideally the research team would like to have developed a much finer gradation of expected risk as part of the modelling of decision response surfaces. The original intention of the research had been to use stated preference techniques to investigate this phenomena (Fishbein and Ajzen, 1975; Towriss, 1984). The pilot activity on the interviews soon showed that this would be impracticable within the time scale and resources of the project and given the particular culture of farmers in the Argolid. Therefore, a simplified version of that approach was adopted which asked the respondent to identify the threat, its perceived magnitude and what responses are most likely. Will this have any effect on the way you farm this crop?: Part 3: Price How much did you receive per kilo for this (main) crop last year?: This is an essential piece of information in order to undertake the economic component of the crop decision model (Chapter 9). In 1994 do you expect the price to be: Much higher; Higher; The same; Lower; Much lower? This question again refers to change that is outside the control of the farmer, this time in revenue. It is an important factor in the model of farmers decision making about choice of crops and level of inputs. The five point range provides a limited representation of risk. Given this change what would you do about the following? Fertilisers; Labour; Pesticides; Choice of crop; Water use; Other responses. This question investigates any potential changes in input costs, and thereby net income, for a given crop and the possibility of adopting a more radical response such as changing crops altogether, or ceasing to farm. At what price per kilo would you be prepared to continue growing this crop to cover costs?:


67

This question provides data on two aspects of interest. Indirectly it enables an estimate of what farmers perceive their costs to be. There is potentially a divergence between real total costs and the perception of them given the possibility of hidden labour inputs (i.e. unpaid family contribution) not being considered at full economic cost or capital costs being written off. In addition it is intended to give a distribution of price at which farmers would stay with a given main crop. At what price would you stop growing this crop?: The converse part of the price/crop choice issue is the price at which farmers would cease to grow their current main crop. It should not be assumed that the “continuation” and “change” prices will be the same since the change activity is disruptive and resource consuming. Farmers were also asked whether they would be prepared to uproot their main crop, and if so how much they would require, per stremma, in compensation for doing so? This provides a comparison with recent restructuring and uprooting policies and a spatial distribution of possible responses among the farming population (see Chapter 8). In practice respondents were more able to anticipate the price at which they might stop growing a crop than one which would encourage them to continue. This appeared axiomatic because they were actually growing the crop at present. Part 4: Water Costs How much did you spend on water last year?: In 1994 do you expect the amount you spend on water to: Increase a lot; Increase slightly; Stay the same; Decrease a little; Decrease a lot? What would your response to this be? These questions enable the distribution of water costs to be related to water use and the perceived changes in expenditure on water to be anticipated alongside the range of responses to this change. The following questions were only directed at respondents who used at least some rationed water (i.e. Canal water, community bore-holes). In 1994 do you expect the amount of water that you are allowed from rationed sources to: Increase a lot; Increase slightly; Stay the same; Decrease a little; Decrease a lot? What would be your response to this? The long term rationing of water, whether free or priced is obviously of considerable relevance to farmer crop choice and the modelling of it. Again the perception of the probability of change, this time in quantity, is sought. When you started using your existing water sources what was the level of salt? (ppm) The yield and quality of citrus crops is reduced as the salt content in irrigation water rises above about 450 parts per million (ppm). This information, along with the dates when bore-holes were first drilled enables some of the spatial and temporal distribution of salinisation to be obtained. In 1994 salinisation was the main concern in many parts of the central Argolid Plain where salt content could exceed 9000 ppm. Farmers generally had good factual knowledge about the level of salt in their irrigation water and an overall picture of salt distribution was sought from this information.

68


In 1994 do you expect the level of salt in these water sources to: Increase a lot; Increase slightly; Stay the same; Decrease a little; Decrease a lot What would your response to this be? This again deals with the perceptions of farmers about the probability of change in a resource over which, as individuals, they have little control. What level of salt do you consider to be dangerous to your crop (ppm)? Although the effect of salinisation on plants and trees is well understood scientifically, this may not be the perceptual basis on which farmers themselves respond. In practice, however, farmers appeared well informed, both about salt levels, and the likely impact upon their crops. The information collected from the two hundred interviews in phase two was integrated with the secondary data on land cover, topography, climate, soils etc. to provide an overview of the Argolid Valley and also to support an understanding of the variation within the case study area. Data was not uniformly distributed across the area and the appropriate unit of analysis for the modelling representation could not be at the farm level. Therefore some aggregation or zoning had to be undertaken which retained the integrity of the data but was representative of the differences across the area. The zoning structure adopted for this analysis is discussed in chapter six following a brief introduction to the role of systems thinking and modelling within policy relevant method.

5. COMPLEXITY, SYSTEMS AND MODELS Paul Jeffrey, Roger Seaton and Mark Lemon

The previous chapters have introduced the concept of policy relevant research and the need for integrative method. They have also carried with them an explicit message that the world is complex and as such prediction and thereby planning are risky ventures. This questions the adoption of reductionist approaches which ‘compartmentalise’ processes and then only respond to change within the confines of those often arbitrary boundaries. An approach which Von Foerster (1977), with tongue firmly in cheek, felt was sure to lead to fame and success. Equally, however, it has been argued that this complexity does not absolve of us of a responsibility to manage change (Spedding, 1979) and to fall into a position of fatalistic indifference. The ability to see the whole picture prior to some form of simplification or reductionism and then to attempt to anticipate possible futures is fundamental to adaptive change management and draws upon interpretive and representational skills. This chapter will provide an elementary insight into how systems thinking and modelling tools can support that process. In their paper on a systems approach to failures Fortune and Peters (1990) state that most major failures arise not from simple unknown causes, but from highly complex human activity systems containing large numbers of interconnected subsystems and components, (p. 384) To take one example from the current study, the apparent causal relationship between the excessive use of irrigated water and water salinity and depletion will be seen as fundamental. The relationship, however, provides us with minimal insight into the range of processes that influence the need for, and use of, irrigation water many of which have nothing to do with the physiological requirements of the crops grown. Indeed it is the locking in to such a reductionist interpretation that can result in technological spirals—in the first place to supply water and subsequently to support remedial activity. The role played independently, and collectively, by price support for crops, poor information about viable alternatives and the attitudes of farmers to changes in practice are all influential and in their turn subject to multiple influences. There is a vast literature relating to systems thinking which has developed, primarily out of the natural sciences, and subsequently been adopted and adapted by a range of human based disciplines (i.e. human geography, psychology) and under a number of headings (i.e. operational research, cybernetics and soft systems methodology). Excellent introductions to this approach can be found (Open University, 1983; Wilson, 1984) and it would be pointless to attempt to replicate these here. Similarly, in the context of sustainable development the reader can be directed to texts that have

70


drawn from systems theory and complex systems theory in particular (Clayton and Radcliffe, 1996; Allen, 1997). Areas of social science (i.e. ethnography) have implicitly held the ‘complex’ nature of what they are studying as central to their approach and this can result in insightful description but create problems for comparative analysis. Alternatively attempts to classify the ‘local’ have focused on the average or the norm whereas it is argued that the driver for change, or the creative energy for it, appears at the margins. In other words it is the diversity at the local level that we need to investigate. A simple example is the importance attached to the ‘floating’ voter as the most important determinant in the outcome of an election. Such a group are by their very nature difficult to classify and as such a more complicated qualitative understanding would be sought prior to any grouping or collective prediction. Recognising difference rather than searching for commonality should form the basis of any investigation of human systems. As has been suggested this does not remove the need for classifications, models and frameworks but puts them firmly behind the exploration of difference. Therefore as one moves away from the ‘local’, and particularly the individual, more abstract analytical tools need to be employed to provide a comparative framework. For example the use of information technology by an Argolid farmer and the improved access to Athens as a result of the improved road system could be described in terms of their literal impact— i.e. using the technology for improved information about markets, agronomy etc. and the road for quicker access to the Athens market. However, at a higher level of abstraction they may both refer to information transfer or to the marketing of crops and as such become comparable but not the same. As will be seen in chapter eight, such local differences are important because they underpin the development of a framework (i.e. for crop choice) and can then be considered in the context of that framework. When such a framework has multiple attributes i.e. markets, information, topography then it is likely that these will have a greater or lesser importance in a particular location. For example, the influence of slope on crop choice is less relevant in the Argolid Plain than in the surrounding foothills. Systems attributes such as slope and soil do lend themselves to a more systematic and mechanistic interpretation i.e. certain crops will not grow in some soils and large scale technologies cannot be used on inaccessible terrain. Whether the farmer can afford the technology, is competent to use it, or has ‘Luddite’ sympathies can only be assessed by a clearer understanding of what is perceived at the local level. Inevitably for a systemic overview to be achieved, generalisations will have to be made but these must be based upon this understanding rather than the rationale of the scientific agenda. What appears ‘rationale’ or efficient to the scientist may be incomprehensible to certain actors for reasons that may not enter into the scientific specification of an issue. In order to be policy relevant, therefore, issues must be locally defined and judgment made on the basis of how those local interpretations differ, and why they do so. The reason for this reiteration of many of the points made elsewhere is that while a complex systems view of the world is advocated the methodologies that emerge from this can be equally top down and subject to a technical perspective (Linstone, 1981) as those reductionist and mechanistic approaches they are so critical of. What we are advocating here, and moving slowly towards in the work, is a combination of the qualitative with the complex and as such the integrated method introduced in chapter two is based upon the qualitative exploration of local difference.

COMPLEXITY, SYSTEMS AND MODELS

71

AN INTRODUCTION TO SYSTEMS CONCEPTS It is not the purpose of this text to provide a detailed introduction into an area with a long, diverse and sometimes contradictory epistemology. Some of the basic characteristics of systems and systems thinking will, however, be briefly considered and then discussed in the context of the models or conceptual devices that have been developed to represent a ‘simplified’ version of them. Clark, PerezTrejo and Allen (1995) put forward a number of points to summarise their ‘view’ of systems. • Systems are made up of interconnected elements to constitute a ‘whole’ that is different, if not greater than the sum of its parts.1 The boundaries of a system separate it from the external environment and it is the need to have sufficient redundancy or adaptive capacity within those boundaries that determines the system’s ability to respond to changes from outside (requisite variety). For example the information present within a system may only support a limited number of responses. In order to extend this capability, information will have to be made available from outside, possibly through contact with secondary connections or ‘weak ties’ (Granovetter, 1973). The example of information transfer also exemplifies the futility of ‘closing’ off human systems in a world which continually impinges upon even geographically isolated social systems (i.e. through the market and the creation of ‘need’, competition for the use of raw materials, or various forms of missionary zeal). Once this relationship or interaction is established it is likely that the subsequent responses to change will also lie outside of the ‘closed’ system, thereby making it dependent and unsustainable in isolation. • A number of distinguishing features can therefore distinguish living from non-living systems. They are open rather than closed and their component elements manifest organisational features while being affected by their participation in the behaviour of the system (feedback). This behaviour in turn transforms inputs into outputs and the system as a whole demonstrates varying resilience to external perturbation. The degree of such resilience will invariably be influenced by the amount of diversity within the system. This diversity is commonly linked to the range of species within a system, however, in the human context the variety of individual characteristics is also an indicator of a systems propensity to adapt and respond to uncertainty and conditions of surprise (Lemon and Scamans, 1997). • The complicated relationships between populations and their environment do not conform to mechanical laws but result in the emergence of new systemic forms under conditions of uncertainty. This is a fundamental characteristic which questions the adoption of end-state planning approaches and highlights the importance of exploration to suggest a range of possible futures against which a variety of adaptive responses might be considered. • The adoption of mechanistic reasoning for understanding living systems fails to account for the non-average behaviour of local actors. Indeed it is the existence of micro-diversity2 at the local level which is both the basis of creative behaviour that drives systems (Allen, 1997) and the source of much of the uncertainty which restricts our ability to manage, let alone control, them.

1 We are grateful to Tony Wright for warning against the tendency to consider the whole as greater than the sum of its parts rather than as something which is qualitatively different. 2 Local in this context is not restricted to the lowest hierarchical level but is applicable to a range of actors operating within the same environment (i.e. government ministers within a cabinet as well as farmers within a village).

72


This point is also pertinent in the context of the methodology described here in that it requires an understanding of how populations differ before any attempt is made to move towards a typology based upon common characteristics. If we build upon these points towards the general premise that systems thinking is concerned with establishing which components interact collectively in response to external pressure then two things require consideration. Firstly, what are the relevant components and how do they interact and secondly what is their collective form? This links the reductionist approach to science which focuses upon the nature of identifiable and measurable relationships and seeks ‘rules’ to express them with an ‘holistic’ approach which concentrates upon the whole as being qualitatively different to the sum of its parts (Glaeser, 1988). While it may be possible to bound (closed) mechanical systems and to a lesser extent bio-physical ones they seldom operate or occur in isolation from human influence. This returns us to the methodological discussion raised in chapter three whereby the need to establish an overview of the whole which does not treat bio-physical and anthropogenic systems as independent and discrete was put forward. Underpinning this ‘systems’ approach was the argument that there was no ‘objective’ reality and as such the multiple interpretations attributed to it had to be taken into account because they provided an improved insight into how different actors and agencies might respond to uncertainty and change. There is therefore a need to move between the specialist understanding of ‘bounded’ systems and the ‘holistic’ interpretation which is defined by the ‘system of interest’ (Lemon and Longhurst, 1996; Slocombe, 1990). Indeed it is argued in the context of the policy relevant research described in this text that only when a system’s coherence is understood can the partial areas or sub-systems be investigated (Blatsou, 1996). Furthermore this raises one other point of interest which distinguishes the work, as it stands, from the claims made for other areas of systems methodology—particularly action research. Wilson (1984) sees action research as a progression from the existence of a real world problem situation through the development of ways to describe this situation to the establishment of problem solving methodologies which are tested and then applied in the real world, thereby affecting change. The clarification about what constitutes a problem or system of interest is particularly complicated even when there is a tangible issue such as the degradation of natural resources. This takes us back to multiple and often competing interpretations and while not denying the ultimate aim of this form of action research, i.e. to intervene, such intervention must be cognizant of this variety and as such sympathetic to policy formation which incorporates, often inconsistent, information from the bottom up. To date it is the first stage of Wilson’s model leading to improved ways of describing real world situations that has been the focus of this work and is described in the following chapters. This refers to the establishment of ways to interpret, describe and present the relationship between agricultural production and the state of natural resources—in other words how to model them. Before embarking upon this discussion it will be useful to present some of the key concepts within systems thinking (Table 5–1). Although not all of these are explicitly referred to within the text their relevance should become apparent to the reader. What is a Model and Why Do We Use Them? Integrative method assumes the existence of multiple representations, or models, of a system of interest. It has been suggested that different actors have their own interpretation of how a process evolves and what are considered to be the salient parts of it. The ability to understand how issues


73

Table 5–1 A glossary of central systems concepts (see Clayton and Radcliffe, 1996; Open University, 1983).

are defined and described by stakeholders, including the scientific community, therefore draws upon their own representation or models. These interpretations are subsequently interpreted and (re) presented through the development of additional conceptual and mathematical models by the research scientists. It is one aim of policy relevant method to use these models for the exploration of possible futures under varying conditions and thereby to facilitate improved dialogue between stakeholders. The rest of this chapter will expand upon the potential of models for moving towards this aim. A model is a conceptual device for simplifying or clarifying the real world, however the term can be used in a number of ways. As a noun it implies the representation of some other object or phenomenon, e.g. ‘they built a model of the empire state building from matchsticks. As an adjective the term ‘model’ means a degree of perfection, e.g. ‘she was a model student’. Alternatively, in the guise of a verb, ‘model’ signifies demonstration, e.g. ‘he will model the latest Paris fashions.’ The common attribute which all three uses of the word have is the implication of something which is an abstract (simplified, stylised, idealised etc.) version of another entity. They are ways of describing and representing real world situations. We may, therefore, describe models as analogous instances of the genuine article; a facsimile or an example. Such a description is remarkably broad and subjective. Indeed, it is so inclusive as to be of limited use as a basis for identifying and classifying models. As an extreme example, adherents of the phenomenological school of philosophy would argue that everything we sense is a model because we do not directly experience the world around us (i.e. we move through the physical environment rather than communicate with it). Certainly, the impressions which we call up in our brains when we think are electro-chemical representations of reality. Even if we reject this view of

74


the nature of experience, it is possible to make a case for models being ubiquitous. One only requires a fertile imagination to defend a tree as a model of an office block or beer mats and ash trays to represent the idiosyncratic interpretation of the off-side rule by a linesman. These seemingly naive examples do however demonstrate the universality of models. We use models intuitively all the time. Having suggested that models are perhaps best understood as subjectively defined representations of other things and thereby almost impossible to distinguish from any other phenomena, it might be constructive to investigate their use. Surprisingly enough, the characteristic of models that cause difficulties with regard to deriving an acceptable definition, is precisely the attribute which marks them out as useful tools both in conceptual and practical terms. They are a representation of the real world, and not the real thing, and as such we can manipulate them for a range of purposes. It is important, however, to remember that while models are devices to help us understand the real world more clearly they are also part of that world both through their physical—or cognitive presence and more significantly through the impacts which arise from their development and use. In summary, models are of use because they allow us to: • Simplify—Models provide an opportunity to strip down the complexity of a real world system and construct a representation which is uncomplicated and hopefully more intelligible and easier to understand. The selection of which dimensions of the system to omit or generalise (selection and resolution) for the purposes of the model will clearly influence its usefulness. • Conceptualise—In simple parlance, models help us to get our heads around an issue or problem. Through simplification we are able to picture and perhaps manipulate elements of the system without having to deal with the complexity or size of the whole. • Structure—A model can support the process of structuring a system. When we talk about ‘structuring an argument’ we are referring to the use of a mental model. • Experiment and analyse—Models allows us to artificially run the system of interest into the future, or recreate its past, changing key parameters and monitoring its behaviour. Because it is not the real thing we can use a model as a test bed for experimenting with changes in structure and process. • Test theories—The use of models in science is strongly (though not uniquely) linked to the development of theories. Indeed a theory is a type of model in that it is a representation (albeit a hypothesised one) of some real system. • Communicate—Models are used extensively for communication between individuals and between institutions and individuals, they provide a framework for narration, discussion and the diffusion of information. • Plan ahead—By projecting models through time we can both imagine the future (desirable or otherwise) and anticipate emergent problems/opportunities.

If these are some useful functions of models, how are we to assess the suitability of any single model to a particular situation? The answer to this requires two issues to be considered. Firstly, what is the intended use of the model and secondly what level of abstraction between the model and the modelled is appropriate? Although there are models around us all the time, we do not necessarily recognise them as such or construct them without reason and purpose. For example we would (hopefully) adopt different approaches for simulating the dynamics of predator/prey relationships to


75

a class of five year-olds to one of undergraduates. The intended use of the model will therefore dictate a subset of appropriate models and central to this is a clear understanding about the nature of the ‘audience(s)’ for it (see Stubbs and Lemon, 1996; 1997). Audiences for models inevitably come with their own interpretations of a system and with their own agenda for their use. This is a fundamental consideration in ‘policy relevant’ work where it is the intention that the model will be incorporated into the policy process. It is here that the reservations expressed above about action research become particularly pertinent. For example work may be commissioned and models sponsored to reinforce a specific political stance, similarly the technical nature of a model may make it inaccessible to those who may wish to base decisions upon it and it is this which is of central concern. Models, particularly computer based ones that are invariably seen as decision support tools, are for exploration and explanation. While they are, on occasions, suitable for prediction this capability should not be assumed by decision makers. To return from the caveat of concern about the use of models we will briefly focus upon how that use is matched to an appropriate level of abstraction. The intended use of the model will help determine which processes, structures etc. are represented and with what level of resolution. For example, if we were seeking to use a process model to test a theory concerning the social behaviour of bees around a hive, we might be concerned to ensure that the frequency and nature of inter-bee encounters were represented as realistically as possible. Conversely, we would probably be ambivalent as to whether the size of the bees was represented at all. It is the nature of a model that it is an imperfect representation of the real thing. The ideal model of say the Taj Mahal would be the Taj Mahal itself.3 Any model will be imperfect in some way, if only because it cannot occupy the same spatio-temporal location or indeed cultural position. Hence, model selection and/or construction should begin with the question: What are the nature of the variances between the model and the modelled and how are these significant with regard to the model’s intended use? Models and science Models have been widely exploited within the realm of scientific enquiry. In fact, scientific knowledge itself has been described as ‘a sequence of abstract models’ (Rosenbleuth and Wiener, 1945). A significant aspect of this relationship has been the almost unnerving reliance upon mathematics as the basis for modelling physical phenomena. Although nobody is quite sure why mathematics should be so appropriate, mathematical models are often able to describe in varying degrees of detail the relationships and processes which characterise our natural world. Two points of caution are warranted however. Firstly, the utility of models within the sphere of scientific enquiry is not limited to the construction of symbolic or physical representations which exactly or closely mimic the real world. Such a function is a particular, albeit very powerful, application for models, and reflects their role in supporting the construction and testing of theories. Of equal value with regard to scientific enquiry is a model’s capacity to communicate or structure or simplify. Secondly, because models are, at root, neither ‘sheer fictions’ nor ‘adequate descriptions’ but analogies, they will eventually break down (Caldin, 1949). Although it is founded on credible models, science has been bedeviled by misunderstandings as to the nature of models as analogies (Jeffrey, 1996). For example, Maxwell’s equations concerning the behaviour of light are similar in form to those which represent the wave motion of a stretched string. Some nineteenth century physicists assumed that analogous equations correspond to analogous realities and were subsequently confounded by later empirical work which found that even if there was an ‘ether’ (the 3

Lewis Carroll cleverly demonstrates this point concerning levels of abstraction in his book, Sylvie and Bruno (1988—ch18).

76


supposed medium through which light travels), it did not have the mechanical properties of matter as inferred by the equations. Similarly, the question as to how science could come up with two incompatible views of the nature of light were misplaced. The competing wave and particle models are not differing views about the nature of light but about analogies to describe its behaviour. A TYPOLOGY OF MODELS Attempts at a formal classification of model types within a scientific framework can be traced back to Rosenbleuth and Wiener (1945). Emphasising the role of abstraction, they suggested a broad categorisation of models into ‘material’ and ‘formal’ types. Material models were viewed as representations of a complex system by one that is assumed to be simpler and has some properties that are similar to those selected for study in the original complex system. They defined a formal model as a symbolic assertion in logical terms of an idealised, relatively simple, situation which shared the structural properties of the system it was representing. Subsequent authors embellished this framework, adding further classes and sub-classes such as iconic, symbolic, verbal etc. (Tables 5–2 and 5–3.) Table 5–2 A general classification of models (after Mihram and Mihram, 1974; Springer et al., 1965; Churchman et al., 1957).

Such a classification is concerned with the distinctions between different types of models and has value as a framework within which to discuss and debate the role of models. However, of equal value is an insight into how different types of model can be described and examples of where they might be developed and employed (Table 5–3). The models represented in the table are not discrete. It would be perfectly possible, and indeed likely, for conceptual, qualitative and process characteristics to coincide, for example in an attempt to represent the diffusion of information and its subsequent assimilation into knowledge. Central to this classification is the difference between linear and non-linear approaches in which linear relationships are based upon proportional (predictable) interactions between components whereas non linear interactions are not proportional or predictable (Table 5–1). In the context of


Table 5–3 Model types adopted within the scientific community.

77

78


the current work a number of distinctive modelling approaches have been adopted with the stated intention of moving towards integration (see Chapter 2). There are few mathematical models which are genuinely interactive, the tendency is for models to focus on one class of phenomena and to represent others in the boundary conditions. One of the most important contributions attempted by this research project is to demonstrate, in a policy relevant context, what contribution can be made by integrating different modelling formats and in so doing moving towards complex systems models (see Chapter 14). A distinction is proposed between models and method which: • arise primarily from a theoretical research imperative • respond to specific classes of decision issues • focus on the design process.

Theory driven models may be useful in that part of the policy formulation process is concerned with exploration of possible futures. Issue driven models and methods will attempt to facilitate the interaction of qualitative information with quantitative information. They therefore promote methods that exploit knowledge from the social and physical sciences within a policy relevant and policy decision-making framework. Design driven models and methods, particularly when the end product is a physical artifact such as infrastructure, will necessarily be concerned about specific configurations of a system or service over a specific time horizon and will normally use the concept of optimisation in some way. Each type of contribution is needed at an appropriate phase in the policy process. Confusion over the roles and contributions of models and methods along this classification may have led in the past to a failure to adequately exploit or develop them. In the longer term it may be desirable to consider what benefits may be achieved if these activities were better integrated. In policy relevant research it


79

is appropriate to develop the methodologies, models and techniques which attempt to capture all these interactions together. These may be particularly relevant for policy exploration over strategic policy time scales. This involves consideration of how they: • • • •

handle the dynamics of hierarchical spatial structures represent the opportunity space of individuals, households and organisations capture diverse responses to change in accessibility and the transport environment manage boundary conditions, different temporalities and ecological issues.

This conceptual outline of, and argument for, the adoption of policy relevant methods has provided the basis for the work carried out in the Argolid Valley, the activities of which are reported in the following chapters.

80

6. AGRICULTURAL PRODUCTION AND CHANGE Mark Lemon and Nenia Blatsou

It has been seen that significant change has taken place in the landscape of the Argolid Valley over the past forty years, primarily in response to the increased availability of irrigation water. It would be very wrong, however, to think of the Argolid as an homogeneous unit and it is central to this study that an understanding of the degradation process, and any appropriate responses to it, must take into account the social and physical variation within the area. It will be the purpose of this and the following chapters to provide a record of the changes in agricultural production for the area as a whole and the variation within it and the different socio-economic structures and technologies that support, and are determined by, that production. Much of the data used in this section has been obtained from the interviews carried out with farmers throughout the valley in 19931 although additional interviews were undertaken in 1996 and 1997 to update the picture. The background to this field work has already been discussed in Chapter three. The activities were designed to provide information about the perceived responses of farmers to changing agricultural circumstances and different levels of degradation. In order to proceed towards this objective it was necessary to obtain information about current farming behaviour. This provided an extensive data base on agricultural production, the technologies employed to support production and the costs incurred and income generated. This data has provided a comprehensive insight into the spatial and temporal aspects of agricultural activity that is not readily available from the official statistics. This chapter will provide a descriptive representation of the spatio-temporal transformation of agriculture in the Argolid over the past forty years. It will also introduce the economic criteria and the organisational and institutional arrangements that have underpinned these changes. It is important to bear in mind that these are only partial determinants of farmer decision making and a more complete picture cannot emerge until a comprehensive understanding of the socio-cultural aspects of the system is obtained. CROP CHARACTERISTICS AND THE HISTORY OF AGRICULTURAL PRODUCTION The agricultural landscape of the Argolid has changed considerably since the 1920’s with the most rapid period of change being facilitated by the expansion of irrigated land between 1960 and 1990. The movement away from rain fed crops such as cereals and vines towards the irrigated citrus crops was mirrored by a concurrent trend towards the monocropping of citrus in the central valley and the retention of a more diverse agriculture in the periphery. This chapter will examine both the overall trends and the spatial distribution which highlights the variety within them.

82


The period around 1960 can, therefore, be identified as significant in the structuring of agricultural production for the Argolid with a rapid shift towards irrigated agriculture and an accompanying increase in fruit production, in particular citrus crops. In order to appreciate this change it is necessary to examine the history of the main crops, their bio-physical attributes and the socio-economic influences over their adoption and rejection. The crop characteristics will now be considered alongside the historical and spatial emergence of agriculture in the Argolid. The recognition that decisions relating to crops are not determined purely by their bio-physical characteristics but are influenced by the cultural, economic, technological and political conditions that prevail will subsequently form the basis of the the crop choice framework outlined in chapter nine. Table 6–1 shows the change in production that has accompanied, and contributed to, this expansion. This translates into the following percentage change for the main crops (Table 6–2) which indicates a movement, across the area as a whole, away from rain fed crops and towards irrigated agriculture, in particular the production of citrus crops (oranges, mandarins). The exception to this trend has been the expansion in olive trees. Prior to 1990 this occurred primarily in the hills however in the late 1990’s with a combination of improved prices and limited water there has also been an increase in olive trees in the central valley. Table 6–1 Changes in area (hectares) for the main crops grown in the Argolid Valley.

Source: Agricultural University of Athens. Table 6–2 % Crop change in the Argolid Valley (1978–1991).

The following is an example of how one farmer has perceived the changes in agriculture over the past seventy years. The respondent draws upon his own experience and that of his family who have

1

This was estimated at 10% of the farming population.

AGRICULTURAL PRODUCTION AND CHANGE

83

farmed land in Kefalari on the edge of the Argolid Plain since the 1920’s. The example provides a graphic picture of how agriculture has evolved, although as will be seen below there has been noticeable variation within the area, particularly over the past forty years. (See Figure 6–1) The history and characteristics of the main crop changes described above will now be examined in more detail. By considering these together it will be possible to establish a more comprehensive understanding of the background to the spatial distribution of crops that has emerged and the opportunity (possibility) and decision (perceived) space within which farmers are currently operating. These options are not restricted to the bio-physical and economic attributes of a crop, an important point that will be expanded upon over the next few chapters. The sections referring to the characteristics of crops are based upon interviews with local agronomists carried out in 1993 and 1997 and the prices given refer to 1993 unless otherwise stated. Apricots: Crop History The introduction of the apricot variety ‘Early Tiryns’ or Tiryntha coincided with the expansion of oranges in the late 1950’s and early 1960’s. The initial expansion of the crop was however considerably slower than that for oranges. Early Tiryns was a new variety supposedly developed by two Argos based agriculturists—Zervos and Damianakos—the latter of whom established the first apricot seed beds in the village of Koutsopodi. (interview with Farm Guards from Koutsopodi, 1993). When the first crop reached full production in the mid-1960’s the producers were unable to establish an adequate market. This was primarily, because at a local level, the crop was new and untried by the consumer. The response by many farmers to this lack of demand was to uproot their crops, however, two years later a new market appeared with the establishment of factories for fruit processing and some replanting took place. The crop, especially the Tiryntha variety, was subsequently exported to Germany and with increased local consumption became the most profitable in the area. Production tended to be restricted to the peripheral villages whereas oranges expanded across the whole area but were mono-cropped in the plain. This was largely because the Apricot trees prefer soils that are rich in organic content and have good drainage, conditions found on the hills but not in the valley where there was a danger of saturation. With this expansion in the peripheral hillside villages 70% of the apricots were co-cultivated with oranges, for example by alternating trees. This was because the Tirynthas apricot was susceptible to the Sharka virus and farmers did not want to run the risk of losing a complete crop. Similarly they felt more confident about having one crop with a certain market through price support—oranges— and another which had greater earning potential but was free market and therefore less predictable. Co-cultivation was therefore one way of spreading risk. It also meant that the labour for picking could be spread. Apricots, like oranges are not labour intensive but they do require a considerable amount of labour for the picking season. Oranges are picked between November and February whereas early apricots (Tirynthas) are picked in late May and the Bebekou variety in early June. This means that in the villages on the Western periphery, where much of the labour is carried out by the farm household, it is possible to spread this more intensive work between the two periods. Sharka virus The intention to ‘spread’ risk by co-cultivation was justified, and reinforced by the extensive impact of the Sharka virus in 1985. The Tiryntha apricot is particularly vulnerable to the virus although it also can affect the Bebekou variety which is also cultivated in the area. The virus which attacks the tree as well as the crop creates necrotic rings in the crop and affects its taste.

84


There is no way of treating the virus apart from uprooting to protect neighbouring trees. This was the ‘solution’ which was adopted and supported by the European Community.

In 1945, when I was about fifteen years old we had around sixty stremmata. At that time we did not grow any perennials but grew corn, tomatoes, aubergines, beans, brooms, cotton and okra (lady s fingers). We also grew vines for wine and raisins although from what I remember, most of the vines in the area had already been uprooted. My father came here in 1924 and bought this farm as one parcel of one hundred stremmata from an American, or a Greek who had returned from America. He purchased the farm with his brothers and at the time it was cultivated with vines. In the first year they almost covered the cost of buying the farm by selling the crop. They earned 400,000 Drs which I would say is the equivalent of two billion Drs at current prices. At that time the price of meat, which was a luxury, was 0.2 Drs/kg. My uncles uprooted the vines when the market price for grapes dropped. They started growing other crops and planted tomatoes. The profit from one stremma of tomatoes was double that which they could earn from their vines. The first farmers who adopted new crops, such as tomatoes, made lots of money and after this all the farms were planted either with tomatoes or beans. We grew cotton when we were married forty or so years ago. It was planted in May and harvested in September. Cotton was also stopped when it became unprofitable. All the crops, when they were first adopted were profitable but as soon as everybody started growing the crop, the prices dropped. At the time corn and other cereals, which were used for brooms, ceased to be grown because although the people worked hard they were starving. Tomatoes, however, used to give a good profit, even when their price started to fall One stremma of tomatoes would produce five tonnes and the dealer would come and buy from your farm. The tomatoes were sold to the islands. My dealer was from one and got better prices there than other dealers could selling in Athens. He made millions of Drachmas. He would start in Argos and move up to Kefalari buying from whichever farm he considered to have good produce and used to buy sufficient produce to fill a whole boat. The dealers would buy the early tomatoes which were usually the best while the lower quality tomatoes were sold to the factories for canning. The tomato market collapsed about 1950 because tomatoes were planted everywhere, even in flower pots. Millions of tonnes were produced, much of which and could not be absorbed by the market. Very often the excess would be left to rot on the farm and the era of tomatoes was replaced by that of citrus trees. After the Germans went (1945) we started to grow oranges on several parcels. My father already grew some trees, common oranges and about thirty lemon which used to produce lemons during the whole year. They were the good years when the lemon trees were not affected by frost, but things changed and we started having severe frost problems and the lemon trees died. FIGURE 6–1 PERSONAL HISTORY OF CHANGE IN THE AGRICULTURAL LANDSCAPE OF THE ARGOLID VALLEY.

The programmes for uprooting the Tirynthas and Bebekou varieties of apricot in response to Sharka and the subsequent conversion of the cultivated land to other crops (E/89/949 and 2/5/VII. 1989) were introduced in 1989 and continued until 1993. The Community provided 70% of the


85

subsidy and the Greek government paid the remaining 30%. A subsidy of 640 Ecu per stremma was paid for uprooting and 150 Ecu per stremma to replant suggested crops. Additionally the programme provided 510 Ecu/stremma, in one year grants, over a five year period for full-time farmers to grow the suggested crops with the same percentage of the costs being met by the EU and Greek Government as for uprooting. The poor take up of the suggested crops (Pistachio and Clover) provides an excellent example of the need to look beyond economic and bio-physical attributes as the basis of crop decisions. Other reasons cited were that there was no track record of pistachio in particular; the subsidy did not extend until a mature production was ensured and orange trees could be planted with the certainty of a market and knowledge about production. The current situation A number of farmers did not uproot their trees when they were affected by Sharka preferring to continue making an income from them until the tree could no longer produce and needed to be uprooted. Other responses to the virus were to leave land uncultivated and wait for more acceptable alternatives to be suggested by the Greek Service of Agriculture. Where cocultivation had taken place, the Apricot trees were often replaced by orange trees thereby increasing the number of single crop parcels in the periphery. Finally, there have been a number of farmers who have rented their land for the cultivation of tobacco or vegetables. There has also been a tendency to plant more apricots even though in the early 1990’s many of the farmers had originally agreed not to do so for ten years in order to qualify for the uprooting subsidy. A new variety of apricot, “Halko”, was introduced by nursery growers and taken up by farmers in western part of the valley and the foothills around Kiveri and Elliniko. This was, however, considered unproductive and uprooted to be replaced by late apricots which provide a good income and can be marketed via the co-operatives for export, canning and juice and sold to the local market. This has coincided over the last three years with a decline in the price of oranges. The characteristics of each of the main crops in the area have been summarised in order to provide a background to the bio-physical and economic influences upon crop decisions. They also indicate some of the indirect influences over crop choice that will be explored in later chapters e.g. labour and capital requirements and the ability of different types of farmer to respond to these. Summary of apricot characteristics Varieties: Tirynthas, Bebekou, Haciotika, Halko. Density: 50 trees/stremma. Planting: January to March. Soil: Apricots are better in soil with good drainage and rich in organic content. Production time: 3–5 years. Bebekou and Tirynthas 60–80 produce kgs per tree, Haciotika 100–120 per tree. Productive life of tree 25–30 years. Form of produce: The Bebekou apricot are sold fresh, for jam and for canning, Tirynthas are generally sold fresh, and the Haciotika for jam and canning. Shelf life/Storage: They contain little sugar and have a short shelf life of 10–15 days under natural conditions. Pruning time: The Tirynthas variety is pruned at the end of May and one stremma will take approximately seven hours. Ploughing cost: 6,000 drachmas per stremma. Fertilizer cost dm/st: 3,000 drachmas per stremma in fertilizer with an additional 1,000 Drs for labour.

86


Picking time: Tirynthas 20th-30th May and Bebekou 10–15 days later Labour for picking and transport: 5,000 drachmas per stremma Labour drachmas per kilogram: Between ten and twelve Drs/kg with more local, household, labour than for oranges. Short season Irrigation/hills: Weekly from May-July, then every 10–15 days Irrigation/valley: Three times in June, then every 15–20 days Water used: 500 tonnes per stremma (Less with rain, good rain is estimated at 100 tonnes stremma) Crop vulnerability: Tirynthas is very vulnerable to the Sharka virus and Bebekou slightly less. Birds also eat the fruit buds in dry years with no grass. Pesticides: Spray 5–6 times a year against Coryneum, Monilei, Oidium, Anarsia. The crop may need additional sprays when the fruit is vulnerable to fungi and insects i.e. if there is heavy rain when the tree is in blossom, or at harvest time. Co-cultivation: 70% of the 1993 crop was co-cultivated with oranges Price per kilo: Tirynthas are twice as expensive as the other varieties and the difference of a few days between picking, often due to local climatic differences, may result in a price variation of 20–40 drachmas per kilo (i.e. 280 Drs kilo in Ermioni, and 250 kilo in Skafadaki). In 1996 the apricots sold through the co-operatives reached prices of 130 Drs/kg. Subsidy levels: There are no subsidies on any apricots Market: Tirynthas and Bebekou are sold to the local markets; Haciotika and Bebekou are sold for processing and Tirynthas are exported, mainly to Germany. Vines: Crop History Vines are a traditional crop in both the main valley and the foothills, however, biophysical conditions of the latter are suited to the crop which prefers high calcium soils and can be grown on slopes and without irrigation. Vines have been produced for domestic consumption and for sale in the local markets the latter of which became flooded in the 1950’s when much of the production was replaced by the cultivation of vegetables, and in particular tomatoes. The introduction of EU support to uproot the crop (reg. 456/80; 777/85 and 895/85) led to a further reduction in production in the late 1980’s and early 90’s however vines have continued to be grown in the villages on the periphery although largely for domestic consumption. Since 1993 there has been some independent replanting of vines in response to the guarantees of local exporters to provide a market. Vine characteristics Varieties: Table wines—Corinthiaki (red), Sultanina—(white) Victoria (new variety). Density of crop: 300–350 vines per stremma. Planting: February to March, however, it is now possible to plant using plastic bags until May. Soil: Prefer soil to be high in calcium. Production time: Picked between June and September with a production ranging from 600 to 1, 800kgs/stremma dependent upon the variety and the cultivation. Productive life of 30–50 years. Form of produce: 250% more vines are grown for wine than eating (table) grapes in the area. Shelf life/Storage: Victorias can be kept refrigerated for 20 days.


87

Pruning time: Mid January. Also between May and June any additional branches are removed. This is usually undertaken by the farmer and one person can prune more than one stremma per day. Ploughing cost: 3,000 drachmas per stremma once per year. Fertilizer: Ammonia (NH4), Phosphates and Potassium, costing approximately 5,000 drachmas per stremma and usually applied by the farmer. Labour and transport: This is usually undertaken by the farmer however an approximate cost is two drachmas per kilogram. Irrigation: Most cultivations are not irrigated and are suited to mountainous or hillside terrain. Those that are irrigated use sprinkler systems. Crop vulnerability: Phylloxera, Evideim, (insects); Oidium, Botrylis, Islia, Phomopsis (fungi) and Peronosporos (the main threat). Pesticides: Approximately 10,000 drachmas per stremma/year are spent on the purchase and application of pesticides. Co-cultivation: Vines are grown alone. Price per kilo: In 1993 the Victoria vine raised 75 Drs/kilo for exportation. Prices varied considerably because many are sold to the local markets at the going rate which reached 200 Drs/ kilo. Subsidy levels: All subsidies apart from that in support of uprooting have ceased. Market: Most grapes are grown for the internal and local markets, local shops etc. although in recent years some have been exported. Many farms also have their own press which are generally used for producing wines for domestic or local consumption. There are, however, no bottling plants for wine in the Argolid. Olives: Crop History Olives are indigenous to the area and until very recently (two years ago) were only grown as a rainfed crop in the hills around the Argolid Plain. The crop is grown for olive oil and table olives and has been considered by the hill and mountain farmers as a profitable crop since the emergence of a wider market for olive oil and particularly with the arrival of European Union subsidies in the 1980’s. Farmers in these areas tend not to record their olive crops by the stremma but by the number of roots or trees. This is because the trees can be hundreds of years old and therefore are not cultivated in plantations but are managed according to their original location which coincided with very different agricultural practices and land ownership patterns. Indeed in many cases the olive trees may not have been considered a viable crop. A similar situation existed with the grapefruit crop in the 1990’s where the fruit was left on the trees because of the absence of a worthwhile market. (See Figure 6–2) More recently the price of olives, and olive oil, has been very sensitive and between 1992–95 dropped dramatically due to scandals which involved the claiming of prices for top quality oils which had been mixed with oil of inferior quality. The reduced prices were also a response to extensive competition from vegetable oils and from Spanish and Italian olive oil producers. One of the reasons offered for this failure to compete was that although the product itself was considered of high quality it was not matched by adequate marketing capability. This would appear to have improved since 1995 alongside an expansion of the northern European market partly in response to effective marketing of the health benefits of the product. With this increased market potential there has been a recent expansion of irrigated olive cultivation in the Argolid Valley. There has been a preference

88


Figure 6–2 Rain fed olive groves in the north east.

for the Koroneiki variety of olive tree which has a short tree and produces after five years which is considerably earlier than other varieties. Olive characteristics Varieties: Manaki, Latholia and Koreneiki. Density of crop: Approximately ten trees per stremma although the number of roots are usually recorded rather than the number of stremmata covered by the crop. Planting: March to May. Soil: Olives are usually cultivated in soil where citrus crops cannot be grown. This is often poor quality, with limited water, on a slope etc. although the recent planting of olives in the valley has changed this to an extent. Production time: The crop is perennial and trees can live and produce for hundreds of years. Form of produce: Mostly grown for oil in the Argolid valley, a few will be grown for table olives. Shelf life/Storage: Oil can be kept for three to four years, but is usually sold well before then. The oil sold in supermarkets is generally one year old. Pruning: Is carried out after picking in March. 50% of farmers will prune their own trees, however with large farms specialist pruners will be employed at a rate of 8–9,000 drachmas/day. Ploughing: This will only be undertaken if the tractor can reach the crop, i.e. due to the slope. Fertilizer: Around 4–5 kilos is applied per tree equating to approximately 1,600 drachmas per tree/ year. The work is undertaken by the farmer or migrant labour at 3,000 drachmas per day


89

Picking: The crop is harvested between November and February. Most of the farmers do pick their own alongside migrant (often Albanian) labour who are paid around 3,000 drachmas per day. When Greek labour is employed payment usually takes the form of up to 50% of the produce. Irrigation: 80% of the olive trees are not irrigated. When a crop is irrigated it is usually 2–3 times a year with approximately 2–3 tonnes. One kg of olive oil requires very little water whereas the equivalent amount of corn oil requires 40 kgs of water. Crop vulnerability: The crop is vulnerable to the insect Dacus Olea, however, providing a community meets a specified level of production the state sprays the crops and the farmers pay for this service at the oil processing plant according to the production in kilograms. Each community decides whether to spray by air or by hand pump. Co-cultivation: Occasionally fodder plants or cereals, which need less water will be cultivated between the trees. Vetch was previously planted to replace the soil nitrate. Price per kilo: One kilogram of oil can be obtained from 8 kilos of olives. In 1992–93 the price was 550Drs/kilo compared with 1,200Drs/kilo in the summer of 1992. The price rose again in 1995–6 however has dropped back to 800 drachmas per kilo in 1997. Subsidy levels: In 1992 four tonnes of oil received 2,800,000 drachmas in subsidy, (see DeWaal, 1991). The subsidy calculation is complicated but approximates at 120Drs/kilo. Larger producers are paid by the kilo whereas the smaller farmers declare the number of trees they own and are aggregated into zones. The production for each zone is then averaged over three years to calculate the current subsidy. This makes the subsidy sensitive to previous production problems and is difficult for farmers to calculate and plan around. Market: Most of the oil is exported, particularly to Italy, however much is consumed locally with certain communities having a reputation for exceptional oil and eating olives. The remaining local oil is bottled and sold to supermarkets. Tobacco: Crop History Tobacco has a long cultivation history in the Argolid, originally as a rainfed crop in the hills and mountains of the north-east and more recently through irrigation in the main valley. Production of the crop was stimulated in the 1920’s with the introduction of licences for its cultivation. These were introduced by the Greek government in an attempt to encourage mountain farmers to remain in their villages and thereby reduce the high rate of depopulation. However, in the late 1940’s and 1950’s the licence holders began to rent land in the valley and to use their licences for the irrigated cultivation of tobacco. This movement of the licences with their holders meant that the aim to reduce depopulation was not met, indeed the opposite effect was recorded. Advice and information about tobacco is the responsibility of the Organisation of Tobacco, which until the early 1980’s also collected and traded the produce. Currently agreement is reached between the farmers and independent tobacco traders prior to production (ten months), the tobacco is then processed through the Organisations ‘driers’ (Xerantiria) and sold to the dealers. This has meant that the emphasis upon quality which was previously insisted upon by the Organisation of Tobacco has been replaced by one intent on maximizing production and this in turn has reinforced irrigated production methods used in the valley. Therefore under the quality control of the Organisation of Tobacco the crop that was previously grown in the mountains was of higher quality but at lower levels of production. Production in the valley is also easier to co-ordinate with large parcels of land closer together, in the mountains parcels are very small and highly dispersed. For example one

90


Figure 6–3 Drying tobacco in north west foot-hills.

farmer with six stremmata of land had this dispersed over twenty parcels (Interview with agronomist of Organisation of Tobacco, Nafplio, 1996). More recently the production of tobacco has decreased dramatically, primarily because of the implementation of European Union policies which reject the established variety (Mavro) as not being marketable. The Mavro tobacco grown in the Argolid is heavy and as such does not meet the increased demand for a lighter product—particularly from the Northern European market (EU regs. 727/70, 1461/82, 1576/86, 1579/79). Since 1992 quotas have been applied to the amount of tobacco produced by licence holders and there has been a lump sum introduced to encourage the return of licences. In other areas of Greece where tobacco was the main crop alternative varieties were supported by the EU, however, in the Argolid where the crop was not dominant this was not the case. Since 1992 the production of tobacco in the area has reduced from 6,000 to 3,000 tonnes per year. Tobacco characteristics Varieties: Mavro, Tsebelia. Soils and slopes: Tobacco grows better where there is good drainage-hills and can produce good quality leaves in poor soil. Planting: October-November with harvesting between May and early June. After harvesting the crop is dried and is therefore dependant upon good weather. If quick drying is not possible the quality of the crop is markedly reduced.


91

Labour: The tobacco plant is not harvested in one session but is carried out over a thirty day period with labour required as the leaves mature. The lower leaves mature first and are collected with subsequent layers picked as they mature. Most of the labour is provided by the extended family. Irrigation: The crop is mainly rain fed but where it has been irrigated, in the valley, there has also been a more intensive use of fertilizers with the combined result of higher production but thicker leaves and poorer quality. Cost of rented land: 40–60,000 Drs/stremma/year depending upon whether the land has access to irrigation water and its location. Vegetables: Crop History The history of vegetables in the area has been characterised by domestic consumption and the flooding of local markets. The introduction of a new crop (i.e. tomatoes) and its subsequent expansion has invariably resulted in a saturation of the market and a dramatic reduction in the cultivation of that crop. This phenomena is less apparent with perennial crops where the commitment to the crop is greater and the time taken to change, slower. As has already been seen with the historical overview of agriculture in the area there has been considerable variation in the types of vegetables cultivated. Currently the most significant factors are the resistance to salt (e.g. artichokes, lettuce) and the ability to invest in greenhouse infrastructure (e.g. tomatoes). Because of the range of vegetables grown in the area the figures used for this study, and collected from interviews and Agricultural Service census data, only provide a crude insight into vegetable production in the valley. This is exaggerated by the choice of tomatoes, which has considerable price variation, to represent vegetables in the modelling of crop choice (see Chapter ten). What should be borne in mind as a factor common to most vegetable production is that it is labour intensive, both in terms of production and marketing. This is important when considering the variation in cropping options acceptable to different types of farmer and has been a major factor in the replacement of vegetables by citrus crops on many farms. Vegetable characteristics Varieties: Winter vegetables i.e. cabbage, lettuce; summer vegetables i.e. tomatoes, aubergines, cucumbers, beans. Planting: Winter vegetables are planted from June to November. Soil: Rich in organic content—flat Production time: Approximately six months. Outside tomatoes can produce 8–10 tonnes per stremma, those grown in greenhouses approximately 15 tonnes per stremma. Fertilizer: Approximately 50Drs/kilo, more than 500Drs/stremma (for tomatoes). Aubergine and lettuce also use a lot of fertilizers. Picking time: Tomatoes appear in the local markets between March and April. They take 120 days to cultivate and involve considerably more work than the other vegetables. Labour: This is undertaken by farmers and their families, or through collective inter-farm cooperation. Water used: Summer vegetables June-November (i.e. 1992) had to irrigate at 100 tonnes/stremma for 30 irrigations (3,000 tonnes per stremma). Winter vegetables use the same amount of water but do not have to compensate for rain water. Because of this the summer vegetables require more

92


irrigation water. Tomatoes use six kilograms per plant per day at the peak of production. One stremma can supply 1,500 tomato plants using nine tonnes of water per day for two months. The total requirement for one stremma over a one year period is twice that for oranges. Pesticides: Vegetables are sprayed approximately eight times in the six month growing period. Co-cultivation: Rotate winter and summer vegetables. Price per kilo: The price for tomatoes is between 50 and 300 Drs/kilo and can reach 500Drs/kilo before Easter. Subsidy levels: None. Market: For domestic consumption and local markets (Argos, Nafplio and Athens). Cereals: Crop History and Characteristics Cereals have traditionally been grown for domestic consumption and the small parcels of land and their inaccessibility, often in the foothills, has meant that it is difficult to use modern technologies such as harvesters. It has also meant that insufficient quantities have been grown by individual farmers to enter a wider market without more comprehensive forms of co-operative structure similar to those operated for citrus and olives. With the expansion of irrigation some of the land previously used for growing cereals has been taken over by citrus cultivations. Varieties: Hard and soft wheat, corn. Density of crop: Between 250–400kgs/stremma. With extremes ranging from 500kg/st to 0kg/st if there is no rain. Planting: Ten kilograms of seed is used per stremma and is sown with the arrival of the first rain between September and to October. Soil: Wherever there is empty land. Production time: Six months from between October-December to June-July. Form of produce: Home based products, bread. Shelf life/Storage: Seeds 1–2 years. Ploughing cost: This is usually undertaken by the farmer twice a year. However, it would cost approximately 3,000 drachmas per stremma if contracted out. Fertilizer: Not usually used, however when it is a cost of 5,000Drs/stremma is estimated. Farmers often use the money received from cereal subsidy to pay for fertilizers. Picking time: July. If the crop is harvested with combines these are usually hired from contractors for 10% of the production. Water used: Rain. Pesticides: No spraying. Co-cultivation: Only with olives if suitable. No rotation but stubble is used for sheep grazing between August and November. Price per kilo: Approximately 50Drs/kilo, if sold. Subsidy levels: 4,500 Drs/stremma for growing. Until 1992 only hard wheat had been subsidised however from 1993 all the other cereals received subsidy. Market: Mostly cultivated for their own use, less than 5% is for sale (e.g. to bakers).


93

Lemons: Crop History and Characteristics Lemon trees were the first citrus fruit to be grown commercially in the Argolid. They were introduced in the 1930’s but did not expand until the 1950’s and 60’s when they grew alongside the apricot and orange crops. While lemons were the most profitable citrus crop, with good prices and a high productivity per tree, they were also the most vulnerable and a combination of heavy frost and the Coryphoxera fungi (Deuterophoma tracheiphilla) wiped out the crop area in the late 1970’s. Lemons however remain the primary citrus crop in the Corinth region to the north of the Argolid. Soil: Needs good drainage without heavy soils. Water used: Lemons use similar amounts of water to oranges but this must be of good quality with minimum salt content. Crop vulnerability: Lemons are the most sensitive citrus crop and are particularly vulnerable to salty water, frost and Coryphoxera. Price per kilo: Higher than other citrus crops. Subsidy levels: When exported it has a subsidy, although little is exported from the Argolid. The Corinth area of the Peloponnese is the main lemon growing area in the region. Oranges Common oranges have been grown in the main valley since the beginning of the century and some trees around the village of Laloukas are eighty to ninety years old. Farmers in this area recall stories of a priest bringing cuttings from California and using them to pay off a debt with a local farmer. The variety of orange which was favoured for the rapid expansion from the early 1960’s was the Merlin otherwise known as the Washington Navel, Tibbets Navel or Riverside Navel. The Merlin was brought into Greece in 1924 by the Dendrology Department of the Agricultural University of Athens and was cultivated in their nurseries before being distributed to the citrus producing areas of Greece. The dominance of the Merlin variety means that when farmers are talking about growing orange trees they are invariably refering to it, particularly those who farm in the main valley. Other varieties have since become increasingly common with the expansion of Navelina in the peripheral villages since 1981 and the introduction of Newhall, Salustiana and Valencia varieties in response to European Union restructuring projects in the late 1980’s. The adoption of the latter varieties has been more limited because on one hand the farmers in the main valley did not want to break from the well established Merlin, and on the other, the farmers in the periphery had just started to cultivate Navelinas. The Navelina is a clone of the Merlin with softer skin which produces a crop fifteen to thirty days earlier. This can have a critical influence on the ability to market the crop and thereby the price received with the peripheral farmers having an increasing influence. This trend has been reinforced by two additional factors that are crucial to the current shape of agriculture in the Argolid. These are the shift away from price support for oranges and the loss of the Eastern European market (particularly Russia) which was dominant in the late 1980’s and early 1990’s. The political changes in Russia and the introduction of the Food Aid Scheme from the United States have resulted in considerable quantities of Uruguayan oranges being supplied by the US to the new republics. Local exporters also refused to sell oranges on credit to Russia with the result that approximately half of the Russian dealers moved across to other Mediterannean producers such as Turkey and Egypt. This change in the export market has helped to redress the balance towards the more flexible full time farmers in the villages of the periphery who currently farm in a more diverse

94


Figure 6–4 Oranges dominate the agriculture of the plain.

way, are less tied to the production of citrus and who sell their oranges to the local markets. A more comprehensive background to the processes surrounding citrus production will be provided later in this chapter. One other member of the citrus family that has been cultivated extensively has been the Mandarin. Common mandarins were first grown in the early 1960’s in the eastern part of the valley (Assini, Drepanon) and Clementine mandarins were cultivated after 1982 in response to restructuring programmes of the EU (regs. 2511/69 amended as 1204/82). However, when this crop reached full production, no market could be found and further subsidies were introduced for their withdrawal and uprooting. Clementines were again cultivated in 1994 when a market was evident for juicing and for export without reliance upon subsidy or price support. The only current constraint is that the crop is very sensitive to frost and farmers are reluctant to grow it without adequate protection (air mixers, sprinklers). The problem of finding a market for the majority of Common mandarins led to their being uprooted between 1990–1992 (reg. 1204/82). The recent success of the Clementine variety has meant that farmers who did not uproot their Common mandarins are now grafting Clementines to them and entering the market place for Clementines. Orange characteristics Varieties: The most important variety of orange grown in the region are the Merlin (Washington Navels) which are the dominant crop in the valley. Navalinas are the dominant variety in the periphery and tend to be sold to the local markets. Other varieties (Newhall, Sanguinia, Saloustiana,


95

Valencia and common Mandarins including Clementines) are grown in less quantity across the areas. Density of crop: For oranges 40–45 trees per stremma and for mandarins 45–50 trees per stremma. Planting: May to October and when there is no frost in the spring, March to May. Soil: Prefer loam sand that is rich in organic content, this is prevalent in the eastern part of the Argolid Plain where the majority of the monocropping production is centered. Production time: The period leading up to full production is between eight and ten years however a good crop can be produced from four to five years. There is considerable variation in productive potential, from sixty to seventy kilograms per tree in areas with poor soils to one hundred and twenty kilograms per tree with good soils. The production cycle is also important in those areas that are protected from frost, this allows the picking season to be extended and a higher price to be obtained for off peak produce. There is an increase in the level of production until the trees approach fifty years of age, this then levels off until they reach one hundred and declines thereafter. The orange tree has a similar productive life to other citrus fruit. Form of produce: The Common orange and Sanguinia are produced mainly for juice; Valencia, Navalino and New Hall are for consumption as fresh fruit and Merlin are produced both as fresh fruit and for juicing. Shelf life/Storage: After they have been treated with wax at the packing factory the crop can be stored for 3–4 months in coolers. Pruning time: Pruning takes place between April and May on alternate years. It is paid for by tree or stremma and usually takes between two and three days at approximately 9,000 drachmas per day. Ploughing cost: 20% of farmers will not plough but will use chemicals to kill weeds. If the soil is heavy the ploughing will smooth it and allow the fertilizers to penetrate. An estimated cost for ploughing is 6,000 Drs/stremma. Fertilizer dm/st: Approximately 6kg of fertilizer are applied to each tree at a cost of 4,000 Drs per stremma. The labour is usually undertaken by the farmers themselves with the contract costs for hiring a tractor being approximately 500 Drs/stremma. Picking time: Merlin oranges are picked between late November and February where there are no air-mixers to protect against frost. This can be extended to March with mixers and to April if there are sprinklers as well.2 The Navalinos are picked ten to fifteen days earlier than the Merlin and New Hall ten days earlier than this. The Valencia are picked between April and May for exportation and through to October for sale on the internal market. Labour and transport: If a tractor is hired to transport the picked produce from the field to the collection point it will cost 15,000Drs/day. A further one drachma/kilo will be paid for transport from the farm to the juicing factory or to the co-operative for packing. The costs for exportation varies i.e. 35 Drs/kilo to the UK although little was exported in 1993 and 25 Drs/kilo to Eastern Europe. If the crop is to be transported by ship from Nafplio then the fees are met by the dealers and range from ten to fifteen Drs/kilo to Eastern Europe. Labour Drs/kg: Oranges cost approximately 6.5 Drs a kilo to pick (Mandarins are slightly more expensive because the fruit is smaller). In the packing factory labour is seven Drs/kilo and an additional twenty Drs/kilo are paid for factory overheads (electricity and pesticides etc.). Irrigation: In the hills irrigation is only undertaken using sprinklers, 320–400 tonnes/stremma/ year. They start earlier and use more water because of the poor soil. In the valley less than 50% of

96


the crop is irrigated by sprinklers with the remainder using flood irrigation or free flow. Irrigation takes place four to five times per year starting in May and using approximately fifty tonnes/stremma for each irrigation. Sprinkler systems use approximately 400–500kgs/tree which is about 100–150 tonnes/stremma/year although this figure will be increased if the system is also used to protect against frost. Crop vulnerability: The orange crop is vulnerable to frost (particularly the later Merlins), salty water, lack of water and strong winds which can destroy the leaves. It is also vulnerable to insects (Greenfly, Anthnomous, Phyllodetes, Tetranychus, Coccoidea, Alevrodes, Ceratis Capitata or Mediterranean fly and Nematodes which occur in the soil and destroy the roots) and to fungi such as Coryphoxera which appears mainly in lemons but to which oranges are also vulnerable. Other fungi which threaten the crop are Phytophtora which appears in both the fruit and the tree, Penicillum and Botrythis which appear after picking and are protected against in the factory and Adromycosis which can destroy both the roots and the trunk. Pesticides: Weed killers are used at one kg/stremma (1,000Drs/kg). Co-cultivation: Co-cultivation is either by the row or the tree. It was started for two major reasons, to insure against Sharka in apricot plantations and to replace orange trees that had been uprooted because of frost damage. Before Sharka 60–70% was co-cultivation with apricots, now this figure is approximately 30% with the rest monocropping. Price per kilo: In 1993 a low price was 38 drachmas/kilo and a high one 52 drachmas/kilo for export. There was an eight drachma per kilo difference between first and second class oranges. This was dependent upon the size of the oranges, any evidence of disease and how well the co-operative consider the farmer has looked after the crop. A considerable quantity of oranges were sold at relatively low prices for juice in 1996–1997. This was in large part due to the loss of the Eastern European market (Russia) and the volume of oranges that remained unsold. For example 36Drs/Kg could be obtained for juice in 1996 compared to 29Drs/Kg in the following year. The oranges reach their weight early in the season, however, at the end of the season they become lighter through dehydration. In 1993 there was a premium of five drachmas per kilo for those crops picked early and late in the season. Those picked early are lighter, those harvested later are more vulnerable to frost.3 Argolid oranges do tend to be later than elsewhere in the area and this affects their position in the market. Subsidy levels: (1993) EU price support for produce sold to third countries i.e. non community, particularly Eastern Europe (Russia) was between 32–34 Drs/kilo. Support for produce sold within the EU was 11 Drs/kilo. Market: In 1993 the main market was Eastern Europe particularly Russia (Poland, Czech Republic and Hungary have all imported Spanish produce). In 1997 this supported market had diminished considerably (see above) and attention was again focused upon the potential of the local market. There was no export demand in 1993 for Mandarins, internally the crop was sold for 60 Drs/kilo and Clementines for 60–65 Drs/kilo. Legislation: The Restructuring of Citrus fruit production (1981–1990) was designed to replace some citrus products with others (Clementines). Most of the mandarin trees were planted through this project although a decline in the market for the fruit occurred when most trees were still not 2 Both air mixers and artificial rain (sprinklers) are used to protect the crop against frost and can, therefore extend the length of the picking season and thereby increase the price of the crop. 3 Local knowledge cites the 6th and 20th of January as particularly prone to frost; in 1991 there was a sixty day period with frost from mid-December.

Table 6–3 Summary of crop history and attributes.

AGRICULTURAL PRODUCTION AND CHANGE 97

98



99

mature. This led to a further programme for the uprooting of common mandarins but not Clementines 1196/3029 (1990). Comments: All citrus trees suffer from mineral deficiencies, particularly from magnesium, iron and zinc. These are added through spraying or ploughing into the soil. A summary of crop history and the attributes of the main crops grown in the area is provided in Table 6–3. The marketing and organisational aspects of citrus production are dealt with in more detail towards the end of this chapter. Prior to this we will explore the contemporary, and where possible the historical, spatial variation of agricultural land cover in the Argolid Valley and the surrounding area. VARIATION WITHIN THE ARGOLID PRODUCTION SYSTEM An appreciation of the variation within the agricultural production system of the Argolid Valley has two central features for the purpose of this study. Firstly it was important to provide a spatial classification with the required level of definition for the modelling work to be undertaken. Secondly, it is only by having a better understanding of the spatial and temporal differences of the production system, as it has evolved, that we can move towards a picture of those options that are appropriate in the future. The farmers crop decision model discussed in chapters nine and ten is based upon production data (i.e. water and input costs, crop prices, and production levels). Each of these are dependent upon the amount of land that is used to grow different crops and the variation across the area, and the physical variables that differentiate between locations, for example by soil type, slope, micro climate, water quality and availability (see Table 6–4). We have already seen that there are physical constraints on the growing of certain crops in specific areas, i.e. lemon trees on

Figure 6–5 The effect of salinity on mandarin trees.

100


Table 6–4 Sensitivity of citrus yield to water shortfall.

Source: FAO. Table 6–5 Sensitivity of yield to salinity.

Source: FAO.

steep slopes or where there is frost. What is more complicated is the choice between crops that will grow in most areas at different productive capacities i.e. olive trees are less sensitive to salt than orange trees (Table 6–5). (See Figure 6–5) The criteria that have been seen to influence the productive capacity of the crops in the area as a whole were then investigated in terms of their spatial distribution. This required some form of spatial classification and the following zoning procedure was undertaken to establish this. ZONING PROCEDURE Considerable time was spent attempting to re-configure the physical characteristics of the region into a zoned structure using soil, hydrological and topographical data, as well as climate data and crop production figures. The need to adopt contiguous zones for modelling purposes and the different spatial patterns of the various attributes made it particularly difficult to identify zones that retained the integrity of each attribute. What was required therefore, was a pattern or configuration of attributes that best represented the physical and production system of the area. It was felt that the best way to achieve this was to draw upon local expertise. This was initially undertaken through a series of extended interviews4 with local agronomists about the physical and production characteristics of the villages in the region. The interviews also inquired about the type of farming structures that were evident in these villages and their background. This will be discussed in more detail in subsequent chapters. The first task of the interviews was to agree upon the physical and production attributes that could be used to represent the differences within the region and that could be subsequently adapted for the modelling work. For example soil quality, slope or topography and access to water are all indicative of the productive potential for crops in a particular area. This interpretation was then considered alongside a map of the area and each attribute placed on either a three or five point scale (i.e. mountains/hills, foothills/valley, valley; good, medium and poor soils). A matrix was then constructed of these attribute scales alongside each village. Once again it proved impossible to zone through a statistical analysis of the matrix which also retained the need for contiguity between zones. A further process of arbitration was carried out which took the interpretation of the original interviews and, in conjunction with local data, resulted in the zones shown in Figure 6–6 and

4

The interviews lasted a total of approximately eight hours.


101

Figure 6–6 Schematic representation of the allocation of villages to zones. Table 6–6 Summary of the adopted zones (1994).

summarised in Table 6–6. Subsequent analysis of data referring to the spatial variation of cropping and hydrological data within the region broadly supported these zones (see Chapters 7–11) with some reservations about zone one in terms of the variation in topography and its relationship to the depth of accessed water. This was not perceived to be a great problem because the zone was seen to have an independent geology to the rest of the area and thereby a separate single acquifer that was clear of salt. The production characteristics of the zone were, however, consistent with the rest of the area. A number of broad observations can be made from this zoning structure, firstly the problems with salt and frost are predominantly in the eastern part of the valley and secondly the peripheral areas have more variation in their cropping and more difficulty in accessing water. With the exception of zone one the soils are generally poorer as one moves away from the central plain and inevitably the

102


slope increases with the related problems attached to drainage and water conservation. How these factors relate to the type of agriculture and farming structures in the Argolid will now be looked at in more detail. ZONED DISTRIBUTION BY CROP PRODUCTION The zonal structure conveys a general tendency towards monocropping as one moves to the plain and a more diverse production system as one moves away from it. Figure 6–7 is based upon data collected from the Agricultural Service and clearly shows a distinction between the three central zones (two, five and six) and the outer zones.

Figure 6–7 Distribution of main crops (tonnes) by zone.

Zones five and six in particular are dominated by fruit production with more vegetable production than elsewhere in zones six and two. If one looks at the remaining zones it can be seen that a more varied production is apparent with a far higher proportion of dry farming. The cereal production in zone two is higher than would be expected, this is grown in the western areas, away from the plain, where a more mixed farming is undertaken. If we extend this analysis to the spatial distribution of irrigated areas then it is not surprising that the same pattern emerges. It is possible to link the current spatial distribution of irrigated agricultural production to its temporal emergence by considering the approximate dates at which specific irrigated crops were planted across the area. Table 6–7 shows this distribution for oranges and apricots, the main citrus and non citrus irrigated crops. Despite the patchy nature of the data, particularly for apricots,5 this shows firstly the emergence of the crops across the area and secondly a clear difference between the timing for the introduction of oranges and that for apricots.


103

The apricot crop tends to have been introduced on the western side of the valley approximately ten to fifteen years after oranges were planted in the same zones. As has already been discussed some apricots were, however, planted in the area at the same time as oranges were introduced. The crop also requires more local labour both for production and for marketing and as such does not lend itself so easily to part-time farming. This will be returned to in chapters eight and nine when considering the social and economic structure of farming in the area. These factors are significant to the way that agriculture has emerged in the past and more importantly as determinants of future change, particularly in the central plain. There is also a noticeable movement, for the introduction of oranges, away from the central valley to the periphery of the area (1952 and 1958 for zones five and six through to 1969 and 1973 for zones four and seven). Table 6–8 shows the spatial distribution of irrigated land with the central zones having a far higher percentage (in excess of 90% in the case of zone six). This trend is reflected in the water requirements for each zone which is based upon the amount of irrigated crops already under Table 6–7 Mean date for the first planting of oranges and apricots by respondents.

Table 6–8 Percentage of irrigated land.

5

This data refers to the main crops grown by the farmers interviewed, no equivalent data was collected for other crops that they grew.

104


Table 6–9 Water requirements (1,000 cu. mt. year)

Table 6–10 Variation in water cost: income per stremma between zones.

production and is shown in Table 6–9. It is important to bear in mind that in 1994 the ground water in zones five and six was also heavily salinated and most of the water for irrigation was obtained from the Anavalos canal infrastructure. These areas also suffer from more acute frost problems and a considerable amount of additional water is used for frost protection. Using this data it was possible to estimate the relative cost of water by zone in 1993 as a percentage of agricultural income (Table 6–10). This is based upon the total income, and the water cost, per stremma. These figures have been obtained from the prices for each crop in 1993 and the production levels for each zone obtained from official statistics. They have been used in conjunction with the price for water which has been derived from information collected by interview from farmers about bought water and bore hole costs for 1993 (electricity and drilling costs). This calculation is based upon the main crops grown by each of the farmers interviewed and as such is primarily related to irrigated crops. Only three farmers cited olives as their main crop. Of the remaining respondents 168 grew oranges, 19 apricots and two vegetables as their primary crop. Although four respondents cited artichokes as their main crop, however, these farmed in a coastal area to the south east which suffered from acute salination and was not included in the final analysis. Table 6–10 clearly shows the differential cost of irrigated farming throughout the area with zones three, four and seven having a considerably higher water cost to income ratio than the central zones. Two explanations can be put forward for this. Firstly, the central zones have had access to the

6 Water

consumption is based upon estimates of 11,000mt3/stremma for fruit and 22,000 for vegetables.


105

Table 6–11 Mean production (tonnes/str) and price per zone (Drs/kilo).

Anavalos canal infrastructure and as such have not had to drill deeper for ever diminishing water resources, with the accompanying costs. Secondly, the production per unit of land tends to be considerably higher in the centre than in the peripheral zones (due to poorer soil, slope etc.), this is compounded by problems of water depletion. Table 6–11 provides an indication of the variation in orange production with zones two, five and six having a far higher mean tonnage per stremma. Table 6–11 appears to contradict the general explanation that has been put forward for the spatial emergence of irrigated farming in the Argolid. The mean price per kilo of oranges is noticeably lower for zones five and six which contain the main citrus producers and the least diverse farmers who market their produce through the co-operatives at a lower price. This apparent paradox also raises a fundamental question about the future options for the area and the dangers of perpetuating the degradation of the natural resource in the short term. The central zones have a higher proportion of multiple occupation farmers for whom farming is a secondary rather than a primary income. Two important points arise out of this. Firstly, the dependence upon citrus production within the central zones is reinforced by the existence of multiple job holding. This means that a higher proportion of the farming population are unable to farm in a manner which is more time consuming than that required for the production of citrus fruits. Similarly they are less likely to have the time, or the inclination, to sell their produce at local markets, albeit it for higher prices than can be achieved through the supported price system. The choice of other markets for citrus, in conjunction with other crops, is more often preferred by farmers from the peripheral zones and as such helps to explain the price differential mentioned above. This exemplifies a perceived inequity whereby some farmers from the central zones were able to accept decreasing prices. Therefore, because they are less likely to change crops for structural and cultural reasons, they could compensate for lower prices by the higher production per stremma and by drawing upon other, primary sources of income. This was not the case with many full-time farmers, working lower quality land on the periphery of the plain. By 1996 however the position had changed markedly with continued reduction in price support and declining European markets. This has put sufficient pressure on the ability of farmers in the valley to consider the viability of continuing to grow Merlin oranges under price support. Alternatively the farmers on the periphery who have continued to sell smaller quantities to the local markets have maintained higher prices.

106


ORGANISATION OF AGRICULTURAL PRODUCTION IN THE ARGOLID Production and Markets Having looked at the spatial variation of agricultural production in the Argolid we can now examine the marketing structures that are in place to support it. Table 6–3 highlighted a range of markets which exist for the crops produced in the Argolid (i.e. co-operatives, dealers, local markets, home consumption). The selection of market is determined by a number of factors. Firstly the price that can be achieved for a particular crop at a particular time. This is dependent upon such things as the level of price support and the ability to protect a crop against frost thereby ensuring product but also extending the window within which it can be sold. The second factor relates to the social formation of a farming household and its influence on the capacity to spend more rather than less time selling the crop for a higher price. A third factor related to both of these is the propensity towards risk, firstly in the form of uncertainty over climatic conditions and secondly over the fluctuations in the local market price and in the uncertainties of the price support system. The Variation in Markets Adopted by Farmers in the Study Area As a general rule certain crops tend to be sold to specific markets i.e. Merlin oranges through the cooperative system, olives through dealers, vegetables and Navalina oranges at the local markets. If this rule were all encompassing then we could take the spatial distribution of crops outlined above and simply equate that data with the markets used. However, the variation away from this norm provides an insight into the social structures that are in place to support different farming configurations. The time commitment required to market produce locally has already been seen to be far heavier than that which is necessary to sell produce through the co-operative organisations. Table 6–12 and Table 6–13 show the markets used by the farmers across the Argolid Valley, firstly for oranges and secondly for all crops produced. If we take the orange market first, one figure appears to be outside of the model described above for the spatial distribution of crops in the area. This is the high percentage of oranges sold through the co-operative system by farmers in zone four, unlike the other two peripheral zones (seven and three). One explanation for this that will be developed below is that many of the farmers in zone four, particularly around Mykenes, have considerable problems in earning a sufficient income from agriculture. As a result they have obtained work outside of farming and this in turn has restricted the time that they can spend undertaking farm work. It is more convenient for them, therefore, to sell their produce through the co-operative system than it is through the local market or through negotiations directly with dealers. The surprisingly high percentage of orange sales through dealers in zones one, two, three and seven may have a number of possible explanations. Firstly, the farms concerned may be small producers but with a range of other crops and as such in a position to negotiate for a better price than that offered by the co-operatives. Alternatively, as has already been seen, in 1994 they may have not been able to secure a sufficient income through the co-operatives because prices were perceived to have been deflated by producers from the central plain (zones five and six). As has already been discussed the situation was significantly different in 1996–97 with the loss of the Eastern European market and the continual reduction in price support. Finally some of the more distant farms may not have access to a local co-operative and as such may have to sell their produce


107

Table 6–12 Percentage of oranges sold to different markets by zone (1993).

Table 6–13 Percentage of all crops sold to different markets by zone.

to dealers, a procedure which again may now be relatively more profitable than had been the case in the early 1990’s. Even with these new circumstances regarding the profitability of different markets it is possible to identify a considerable difference between the way oranges are sold in zones five and six which is primarily through the co-operatives, from the other zones who also sell to dealers and in the case of zones three and seven to the local markets. It would appear that where there is more variation in crops grown then there is also more variation in the markets used. This extends to those crops that are grown for home consumption such as cereals and vines; zones one and seven have a noticeably higher percentage of crops grown for this purpose. Having looked at the cropping configurations across the Argolid Valley and the variation in the markets used to sell that produce it will be useful to examine in more detail the marketing options that are in place to support agricultural production. Co-operatives The co-operatives have emerged into their present form alongside the increased production of citrus fruits over a thirty year time period. They were originally founded in the late 1940’s to provide production and financial support to farmers. This role has become secondary to the marketing function that they undertake today, particularly in response to the increase in overseas markets. There are about forty co-operatives throughout the Argolid Valley, approximately one per village or community, although some villages may have two or more e.g. Nea Tiryns. The existence of more than one co-operative is not only a response to the size of the village, it is also often an indication of

108


the political divisions within it. Therefore while co-operatives are primarily economic organisations they also operate according to political affiliation and to local family history. The co-operatives market most of the oranges and apricots and a considerable percentage of the mandarin production. Other produce is sold by the farmers themselves, either on the internal market or through dealers. Each co-operative is directed by a council which is elected every three years from its membership which consists of the farmers who use its services. The council generally consists of five people, the Proedros (chair), secretary, treasurer and two other members. It is the role of the council to decide how much is paid per kilo (i.e. what is deducted by the co-operative from the agreed price with the EU) for a crop and there is great disparity both in these decisions, and in the quality of the organisation in place to support them. For example the co-operative of Pyrgella paid only fifteen drachmas per kilo for oranges in 1992 and had not announced the final price or timing of payment by March 1993, neither had they announced the prices for the forthcoming season.7 In contrast the Kourtaki co-operative has a reputation for prompt payment. More generally the situation has become considerably worse over the last three years with many farmers not receiving payment in 1997 for produce of the 1995–96 season. In 1997 produce was sold with no guarantee of when, or if, payment would be made. Prices vary according to the timing of the crop,8 its destination (i.e. local market, export to Northern or Eastern Europe) and the purpose of the market (i.e. juicing, dumping and social withdrawal will all receive lower prices). Other forms of payment may also be negotiated i.e. Fragistas the largest independent exporter has accepted part payment from Russia in timber and the Olympic fruit juice company have paid some farmers with fertiliser. In 1997 Fragistas have refused to accept credit terms with the Russian traders and as a result have lost a considerable amount of that trade to other Mediterannean suppliers. The turnover in two of the largest co-operatives in Inahos and Hera is around 18,000 tonnes each per annum. A relatively small co-operative plant (Perseas in Argos) employs 60 people in the five month season, including one agronomist, one engineer and two managers. Many of the factory work force are women employed in grading and packing. In addition the co-op employs eighty to ninety pickers on a casual basis depending upon the requirements at a particular time. The majority of those employed to work within the plant will be locals, whereas the pickers are likely to be seasonal migrant labour. Because of the dispersed nature of many farms co-operatives have producers throughout the area. Equally it is possible for one farmer to be a member of more than one organisation or to carry produce from a dispersed farm to one co-op. In 1994 the requirements to set up a co-operative9 were that the organisation should have an output of 5,000 tonnes and at least one hundred members.10 If one or other of these criteria are not met a new organisation cannot be formed and an existing one will have to dispand. This qualifies the organisation for price support, the opportunity to bury surplus production and to capital support for the installation of a packing plant, although the buildings in which it is situated may only be rented for four months of the year for packing and distribution.11 If the criteria are not met then it is possible for groups of producers to obtain price support by registering with the Union of Agricultural Co-operatives and to bury excess produce through the Union in Nafplio. Dealers As has already been seen a high proportion of citrus and other crops are sold through dealers, particularly in zones one, two, three and seven. Dealers in fruit are both local and from outside of


109

the area i.e. former Yugoslavia, Poland. Although there are few dealers from the UK, Holland or France, these tend to operate through Greek middlemen who make contact with the farmers and the co-operatives especially those with air mixers (for protection against frost) and a late crop whose produce can generally realise a higher price. In 1997 the crop which was held back to realise these high prices failed to do so because at the peak period of the market there was insufficient sales and a large percentage of the total production remained and competed in the market place. Coincidently there were also unusually heavy late frosts and this also had a negative affect on the quality of the late crop and the price that it could realise. Dealers usually operate through the co-operatives, although some of the larger co-ops also act as dealers. A percentage of the agreed price is paid in advance with the balance being met at a later date, although as has been seen, delays in payment frequently occur and are of variable duration. These delays, along with organisational problems and malpractice have created considerable uncertainty, both for the co-operatives, and even more acutely for the farmers. In 1992 the Perseas co-operative only received twenty two drachmas per kilo for its produce because both produce and money were misplaced during the exportation process. Twenty six drachmas per kilo should have been realised and the overall losses to the co-operative were estimated at sixteen million drachmas. This has profound implications for the cash flow of the producers and causes considerable bitterness amongst the farming community. “There is no money, the dealers do not give the money they should. Anybody can call himself a dealer (by obtaining a licence), they absorb the toil of the farmer and the farmer then has to wait a year for his money. How can he cultivate for the next year when all his obligations (electricity, fertiliser, pesticides etc.) have to be paid in cash”. (Farmer: Pygella) The crux of this particular problem is perceived by farmers to be the payment of price support to exporters who deal directly with the dealers and buyers (middlemen) rather than with the producers. As a result, one of the solutions to the marketing problems is seen to be the establishment of direct contacts between the exporters and the co-operatives or producers. A constraint upon the realisation of this solution is the low level of formal education of many farmers, a factor that will be returned to later. One other perceived effect of this loss of income to the ‘middlemen’ is that it has not only contributed to a squeezing of prices but also to a focus on increased production and to a relaxing of

7

Prices are agreed between the co-operative and members at the outset of each year, these vary considerably between co-ops and between members of the same organisation, depending upon location, frost protection etc. 8In 1997 two types of price support existed. Type A (prokathorismenes—‘prefixed’) is an export agreement whereby the co-operative agrees a production of x tonnes at prefixed prices with the EU. In February 1997 this was 37 Drs/kilo to Russia and 27 Drs/kilo for the rest of Eastern Europe. Type B (annagelias—‘varying’) applies to the remaining production and varies at figures below Type A support. In April 1997 there was no Type B support for Merlin oranges but there was support for the later Valencia variety in May and June. 9 Reg. 1035/72. The programme for the establishment and working of groups of fresh vegetable and fruit producers. This is the legislation upon which the co-operatives are based and which determines price support and excess production quotas. 10 This varies between Prefectures, Corinth only requires fifty members and 1,500 tonnes of produce to qualify. 11 Only five or six co-operatives in the area own their own premises.

110


quality standards. This has led to the exportation of poorer quality produce, with a loss of confidence by the existing market. This has already happened with the olive crop through a number of scandals concerning the mixing of low and high grade oils. Furthermore, in 1996–97 a number of orange shipments were returned due to the poor quality of their produce. Local Markets The role played by internal markets has been seen to be more significant in those areas that have a greater variation in the crops that they produce (zones one, two, three and seven) than those which concentrate primarily on citrus production (zones five and six). Farmers from Skafidaki (in zone one) sell most of their produce, except oranges, at market and only ten out of seventy families from Statheika (in zone three) do not sell some of their produce in Athens. The benefits of selling to the local market are the realisation of immediate income, higher prices and the opportunity to avoid tax. While vegetables and other crops are sold in the local markets more than citrus fruits, when these are sold the same benefits accrue. The risks of selling citrus in this way have previously been greater than if they were sold through the co-operatives under price support, because there was no guaranteed price. In 1997 however, uncertainty surrounding the payment mechanisms and the level of price support reversed this process with the local market offering more security. Of course this does not alter the vulnerability to inclement weather and the benefits to those who work the land that has good water and rare frosts. Similarly there is more flexibility for those who are already attending local markets to sell other produce. These points are reinforced by the Secretary of the Community in Skafidaki12 “80% of the money is earned in this way (at the local markets), there are no families that do not have someone who goes to market. This is an early producing area with no frost problems (i.e. low risk, high price)”. Selling through the local markets commits the farmer for the time that he/she is at market, this means that the rest of the family need to work the farm in their absence. This is only possible if there is some form of support network, either within the family or between neighbours. Such a format might consist of the farmer attending market, his partner and children working the farm and parents looking after the youngest generation. Attending market, therefore involves long days13 that are greatly extended by the need to arrive early to obtain a pitch. It is this latter factor that dissuades many farmers from selling through the local markets. The Ministry of Agriculture issues licences for market pitches and there is considerable competition for them with some ill feeling about the fairness of the allocation, particularly with the number of licences that are issued to traders rather than farmers. “I need the money from the market but cannot be bothered to go and fight for a pitch. They kill each other for a seat. If I had a (guaranteed) seat once a week then I would go”. The choice of market and the social structures necessary to support that choice will be discussed further in chapters eight and nine. It is one indicator that will be seen to differentiate between those who are perceived as ‘real farmers’ and those who are not. This distinction is not solely determined by economics but provides important insights into the class and status of agricultural work as


111

opposed to land ownership. Many youngsters reject life on the farm for sedentary work and a lower income. This is less common in those communities that sell to the local market, a point made by one farmer from Skafidaki. “If my child studies and finds a job with 15,000 drachmas a month (£450), this is the money that I can get by attending market just once”.

OTHER MARKET SUPPORT If 30% of produce is exported by any operation (a co-operative or teams of producers) then any excess production of grade one produce can be taken out of the market and buried.14 If it is not possible to export the produce then no restriction exists on the amount dumped. From 1983 until 1991 it was possible to bury oranges, common mandarins, apricots and lemons although the latter was not produced in sufficient quantity in the Argolid Valley. From 1993 this facility has been restricted to Navel (Merlin) oranges and a figure of two drachmas is deducted from the thirty four drachmas price support to pay for the Havuza or dump. Oranges are checked at the dump by Ministry of Agriculture officials who carry out quality control and sign papers confirming the amount buried.15 These are then sent by the co-operative to the central ministry who process them and pass them on to the European Union. Payment is then made through the Agricultural Bank two to three months later. Until 1996–97 farmers were keen to produce a good quality product and were not sympathetic to burying their produce. The dumping or burying of produce was seen as a pointless exercise that led to a general lowering of standards because producers are less inclined to bother about the quality of a crop that is not going to be consumed. In addition to this, fifty percent of the oranges dumped in the valley were produced by people for whom farming is a secondary occupation. This inevitably meant that more people were chasing the same market, but more importantly those who had other incomes were more able to accept the reduction in price that accompanies this process. However, the general uncertainty and price reduction over the last two to three years has meant that many of these farmers are finally being forced into a position where they will look for any available markets including the subsidised dumping of their produce. However, as demand for this option has increased there has not been a corresponding level of planning by the local of Service of Agriculture as to how this process should be managed. Juicing Between 1992 and 1994 a number of factories for juicing fruit were set up or expanded. These used second class produce which was in excess at the peak of the season, or fruit that had been kept too long, was less juicy and could not be sold fresh. At the time a subsidy of twenty eight drachmas per

12 Community secretaries are state employees responsible for collecting and disseminating information about, and relevant to, agricultural production. 13 Farmers often leave for Athens before four in the morning and return in the early evening.

112


kilo was introduced by the EU for juicing oranges. The fruit was purchased by the factories from the producers at thirty two drachmas a kilo which was the price received at the time by the co-operatives for burying produce—for which there was a two drachmas per kilo charge. In effect therefore, the factory only paid four drachmas a kilo for the fruit to be juiced. There are about eight such factories in the Argolid, none of these are new but all have expanded since the introduction of the subsidy alongside additional support of 60–70% for building and capital costs. The Aigli factory in Argos is the largest and can process 800 tonnes of oranges in a twenty four hour period. During the peak picking season from December to February it will employ approximately eighty people on the shop floor and between five and ten for packaging which continues throughout the year. A considerable amount of orange production was sold for juice over the 1996–97 period. This was due, in part, to the loss the Russian market which resulted in huge amounts of oranges remaining unsold and an accompanying drop in prices (from approximately 46Drs/kg for Merlins in 1995 to 36Drs in 1996 and 29Drs/kg in 1997). There was also concern about underpayment for produce whereby the EU provides a 29 Drs/kg subsidy to the farmer for oranges sold for juice and the farmer signs a receipt for 44 Drs/kg with the difference to be paid by the exporter. This difference is often not met with perhaps a 7 Drs/kg payment rather than the 15 Drs/kg which was signed for. One other source of price support, instigated by the European Union and the Greek Government, has been the creation of a social market for oranges. This has operated in two main ways, firstly through a programme to provide fruit for families with over four children and secondly through the supply of oranges to schools. In 1997 2,000 of the 7,000 tonnes distributed by the co-operative in Neo Tiryntha went to schools in the north of the country around Thessalonika. Other Influences on the External Market Recent changes in the external market, such as the loss of the Russian market have already been discussed however earlier in the 1990’s there was also considerable fluctuation and uncertainty. In 1992 there was a 40% reduction in oranges exported from Greece with virtually the whole Western European market disappearing. A number of reasons were put forward to explain this change. Firstly, there has been an increase in foreign competition, most noticeably from Spain and Israel, particularly Spain. The Spanish government provided additional support to the standard eleven drachmas provided by the EU.16 As a result it was perceived by Greek producers that they were able to enter the market at lower prices. In addition to this there was some support for the view that the organisation to support the Spanish orange was more efficient than that which was in place for the Greek product. “There are no records but if one looks at the difference in packaging, organisation and marketing between the Greek and the Spanish system it is easy to see why the latter is more competitive.” (Agronomist Service of Agriculture)

14 15

1035/72. To qualify for dumping, oranges must be of good quality.


113

Secondly, before 1991 (when the Eastern Block existed) the EU price support for oranges sold to East and West Europe was similar with approximately five drachmas difference. This difference rose to over twenty drachmas in 1992. Eastern Europe in this context had Third Country or non EU status and qualified for European Union support as did Austria, Switzerland etc. Indeed some produce was transported into these countries, and in so doing qualified for higher levels of support, prior to being taken into EU countries for which there would have been lower supported prices i.e. from Austria into Germany. A third contributory factor was the war and subsequent disruption in former Yugoslavia which meant a loss of land freight traffic and a reliance upon the movement of produce from the port in Nafplio. Dealers from northern Greece had previously purchased Peloponnese oranges, packaged them in the north and then exported them to the EU by road. CONCLUSIONS Two central features have been observed about the changes in agricultural production in the Argolid Valley. Firstly the system as a whole has moved towards mono-culture and overproduction, this has resulted in physical and economic vulnerability and considerable uncertainty. The second feature is that agriculture has developed with minimal planning, or indeed if one considers the external influences upon agricultural production, the opportunity to plan. Three broad conclusions have emerged from this analysis, each one of which will be returned to in the ensuing chapters. 1. Over a fifty year period agricultural production in the Argolid Valley has moved from dry to irrigated farming with no coherent plan about the suitability of specific areas for different crops (i.e. local climate, soil type, water), or the appropriate amounts of individual crops that could be grown in each area. 2. All of the responses for the restructuring of agricultural production have related to the external market conditions and not to the state of local natural resources. For example the system of price support for juicing factories and for the dumping of excess produce both reinforce existing over production. 3. This pressure to over produce has been compounded by inefficient marketing practices, in particular those in which the payment of price support is not made directly to the producer. The ensuing chapters will examine the technical, cultural and structural processes and conditions that have underpinned agricultural change in the area as a whole and the differences that have emerged within it.

16

Price support for produce sold to the EU countries was around eleven drachmas per kilo in 1993 whereas that sold to non EU countreis had risen to thirty four drachmas per kilo.

114

7. TECHNOLOGY AND AGRICULTURAL PRODUCTION IN THE ARGOLID Mark Lemon and Nenia Blatsou

INTRODUCTION Agricultural production in the Argolid has been increasingly determined by the use of technology, firstly to access irrigation water and secondly to protect against the effects of ongoing agricultural practices on the natural environment (i.e. the response to frost). “Machines have entered our lives, without them we would not even have had water to drink, over the last few years.” (Argos Farmer, 1992) Figure 7–1 provides an indication of this trend over the past thirty years and both this chapter and chapter eleven discuss the qualitative and quantitative impact of technology on the water resources of the area. The final section of this chapter will consider the role of the ‘expert’ as one which is based upon a technological perspective and has effectively reinforced existing forms of agricultural production. This perspective has also been frequently adopted by the farming population, for example the Proedros

Figure 7–1 Units of technology per 1000 stremmata of irrigated land (Source: Service of Agriculture).

116


(head) of Mykenes discusses the drilling of bore-holes in 1974 and the cultivation of oranges using this water. The period until 1994 witnessed an increase in the depth of bore holes to 420 metres, as well as profound uncertainty about finding water. “Most of the bore-holes are dry, if we have thirty five then thirty will be dry. We do not speak about real irrigation but about keeping the trees alive while waiting for the rain. Because of our soil we have a very big problem. The only solution is Anavalos”. The Proedros (President) of Nea Tiryntha makes a similar point: “Anavalos (for the transportation of spring water by canal from the sea to the valley) can solve the problem of the Argolid, with freshwater replenishment, providing there is enough rain so that the ppm of chloride does not increase. Anavalos has huge quantities of water so that the whole Argolid valley can be irrigated if the necessary secondary canals are constructed”. Since 1994 there has been an increase in the annual levels of precipitation from approximately 500mm to over 900mm per year in 1996. Recharge techniques in which water from the Kefalari springs is transported by the Anavalos infrastructure and siphoned back into the aquifer have also been introduced more extensively in the valley over the last two year (1996–97). The combination of these two factors have greatly reduced the concern of farmers about the quality and accessibility of irrigation water and focused their attention more directly upon crop price and market uncertainty. This chapter will consider the role of technology in both the degradation and the remediation of water stocks. It is important to recognise that the current reversal in the apparent degradation of aquifers is not necessarily an indicator of long term sustainability. Therefore the background to the situation which arose prior to 1994 remains relevant to understanding how technological spirals in combination with inappropriate economic incentives can force a system along unsustainable paths. Interviews with farmers in 1997 recorded limited concern about the condition of natural resources and it is not unreasonable to assume that with a prolonged period of lower rainfall, and continued emphasis upon intensive irrigated crops, the previous critical position could reoccur. Excessive rainfall has, however, highlighted some concerns about flooding and there is currently support for the building of dams across, and widening of, dry river beds. Once again the response is to turn towards technology. Each of these issues will be explored later in the chapter, the purpose of which is to document the emergence of those technologies that have firstly enabled irrigated agriculture and then sought to protect it. It is important to see the increased use of technology in agricultural production as scale dependent. The adoption of earlier irrigation technologies were at the level of individual farmers with a minimum of technical input, i.e. horse drawn wells. At the other extreme there is the large scale centrally funded Anavalos dam and canal project the output of which is then distributed amongst a proportion of the farming community. This introduces questions of equity and an overtly political component relating to decisions about the course and the use (i.e. for irrigation or recharge) of the canal. In between these two extremes there are various forms of collective organisation that have emerged to enable technical responses to specific problems when they have been out of the reach of individual farmers. For example the collaborative financing of exploratory drilling for water. The background to each of the technologies perceived by farmers as influential to the type of agricultural production undertaken in the Argolid will now be discussed.

TECHNOLOGY AND AGRICULTURAL PRODUCTION

117

BORE-HOLES AND IRRIGATION PUMPS A limited attempt will be made to analyse the spatial and temporal development of irrigation technology in the area It must however, be borne in mind that with the increase in rainfall and recharging activity over the past two years the number of bore-holes drilled has reduced dramatically. This section will be concerned with providing the background to, and the circumstances in which, this expansion (prior to 1994) took place. A fundamental difference between the large scale transportation of water and its subsequent distribution and the private drilling for water is that in the former case it is purchased directly and in the latter the costs are incurred through drilling and pumping. This is an important difference in a farming community which places a high value upon independence and distancing itself from authority (Herzfeld, 1992), albeit in the form of water authorities. Therefore, having already stated that there is a belief in the ability of large scale technology to overcome current water problems there is a concurrent intention to retain autonomy for as long as possible. As a result, it was conceivable that when, and if, the overall water situation improved then the use of Anavalos water would decline and there would be a return to the use of individual bore-holes. In 1997 this appears to have been the case in some parts of the valley. This trend is important not only for understanding the delay in adopting Anavalos in some areas, although the poor quality of its water was undoubtedly a contributory factor as were the restrictions upon its use (i.e. only for irrigation of citrus crops). It also provides an insight into what might constrain various options for the future. It is suggested, therefore that when the choice exists (i.e. access to the infrastructure and good quality ground-water) private bore-holes will invariably be adopted in preference to the Anavalos system.

Figure 7–2 Pump-house for irrigation.

118


Figure 7–3 Chronological development of pump technology. Table 7–1 Changing water pump technology.

The technology required to access water has become considerably more expensive, as well as extensive, over the past thirty to forty years. The history of bore-holes has been based upon the desire for individuals to have access to their own water source. This raises a number of issues that are central to understanding the water crisis prior to 1994. In the 1950’s and early 1960’s water was relatively near the surface in many areas (less than ten metres). The centrifugal pumps (Figure 7–3) that were used did not need to be highly powered and were generally horse driven or used low powered diesel pumps (Table 7–1). Therefore, water could be relatively easily accessed but at a low pressure which made it difficult to irrigate a large area. This obviously prevented an over use of the ground water resources. However the desire of farmers to have their own bore holes, and the ready provision of capital from the Agricultural Bank of Greece in the 1960’s, meant that pumps were installed in large numbers and close proximity. The inherent dangers in this trend were exacerbated by the introduction of Pomones electric pumps which enabled water to be obtained at a greater depth if this was necessary,1 and to irrigate over a far wider area. What emerged therefore was an unregulated and disorganised use of ground water with bore holes in some areas being drilled every five stremmata. The effect of this activity in those areas that were the first to irrigate (zones five, six and two) was to encourage sea water intrusion in the plain and the need to drill deeper towards the foothills. The progressive

1

Pomones pumps are capable of reaching 200 metres, Epovrehio or submarine pumps which have the pump at the base of the bore-hole can reach a depth of 400 metres.


119

salinisation of ground water is described in chapter eleven, what is important in this context is that it led to the introduction of isolation technologies to control the intrusion of salt. Isolation In those areas which suffered from salt intrusion (in the main valley) new bore holes were generally ‘isolated’ at a depth of between fifty and one hundred metres. From forty to sixty metres there is an area with compact, impermeable rock and no water into which an initial drilling is made.2 A second tube is then placed inside this and drilled slightly deeper. This is left for one or two days to see if water infiltrates between the tubes and if it does there is a problem of sea water intrusion. The response to this will be to drill into the second acquifer at 80–100 metres and to repeat the process (Figure 7–4). A number of problems can arise to prevent successful isolation. Firstly, if the original bore is not tested adequately and a pressure difference results from the isolation then sea water intrusion will result. Secondly, each farmer is vulnerable to the activities of his or her neighbour. If they do not isolate effectively, or fail to isolate at all, then it is likely that the ground water will become salinated and this will permeate the bore-hole of the more diligent neighbour. One other problem that was encountered relates to the use of poorly preserved isolation tubes that are eroded by the salt with inevitable results. It is an ongoing problem that no information exists about bore holes that are not working, and/or, have salinated water. This means that it is often difficult to prevent salt penetration of new boreholes. It would therefore, be of value to identify all the existing bore-holes and to close those that are no longer operational, possibly by cementing them up. The magnitude of such a task should not be underestimated with current estimates of 15,000 bore-holes (Agricultural University of Athens) in the area and with a farming population that is often sceptical about providing information to the ‘authorities’.

Figure 7–4 Isolation protection against salt intrusion (Depths used are examples they do not represent the variation in isolations).

One other attempt at improving drilling activity has been through the introduction of monitoring and licensing procedures. Farmers are currently required to obtain a licence from YEB (Land Improvement Service) before they can open a bore-hole. This is reinforced by the electrical (DEH)

120


Figure 7–5 Sinking a bore hole.

service who will only connect power to the pumping system for the bore-hole on the provision of that licence.3 The success of this control is indicative of the extent to which electrical pumps are used. The procedures only relate, however, to current and legitimate holes, they do not inform about, or regulate, the non-operational and illegitimate holes that have been perceived as a major cause of salt intrusion. YEB do not undertake any checks on the hole during or after its drilling. Indeed the only check carried out at present is to enforce legislation which prevents the drilling of bore-holes within fifty meters of each other. In 1989 500 permits were issued for opening new bore-holes and 250 for deepening existing bores. The current law only relates to the surface distance between holes and does not address the underground conditions. As a result it has not been possible to control the real effect of the drilling. This is seen as a fundamental mistake (Association of Agronomists, 1992) and opens up a number of avenues for abuse in which water becomes a part of the land pricing and control mechanism as well as a resource for production. For example, false bore hole heads are sometimes sunk to prevent neighbours from drilling for water within fifty metres and thereby to protect access to it. Similarly

2

These figures obviously vary from area to area depending upon the geological structure, however the principles of isolation are the same. 3

DEH, the public electric company, are responsible for pursing a Programme of Farm Electrification which provides information about bore-hole installation. The cost of agricultural electricity, including non-dwelling buildings is also lower than the standard rate, approximately fifteen drachmas per kilowatt hour compared with twenty five drachmas per kilowatt hour.


121

when a bore hole is sunk in the corner of one piece of land it means that the neighbouring farmer cannot open another within fifty metres. As a result, where there are small parcels of land without bore-holes, it is possible for individuals or groups to cover a neighbours property and to prevent him/ her drilling for water. This not only hinders irrigated production but has a considerable impact upon the value of the adjoining land. It can be one way of acquiring a neighbouring plot at a reduced price because land without water is often at least 50% lower in value than that with it. Apart from the rejected option to move away from irrigated agriculture, the individual responses to water degradation have therefore been in terms of isolation, deepening, opening more wells and the purchase of bought water.4 Each of the first three responses have become progressively more expensive and have increasingly required a collective response from the farming community. The cost of a bore-hole rose sharply in the early 1990’s in response to a rise in overall prices as well the increased depth required. For example a one hundred metre bore-hole drilled in Pyrgella (zone 6) in 1989 cost 3,500,000 drachmas. A 240 metre bore-hole drilled in Argos (zone 2) in the same year cost 6,000,000 drachmas (approximately £18,000).5 It is difficult for small farmers to raise this sort of capital alone particularly as the loan finance that was previously available through the Agricultural Bank is no longer readily available. The water from some of the collaboratively financed wells was sold by YEB and distributed through a village committee whose president was generally an agronomist from the Service of Agriculture. The distribution of the water operated in a similar way to which the Anavalos infrastructure functions today, through the local TOEB,6 and was intended to avoid disputes between farmers. An additional factor emerged alongside the increased costs of bore-holes, particularly when those additional costs related to the uncertainty of finding reliable water. This involved a movement away from the more stochastic approach to finding water and the introduction of various forms of exploratory organisation designed to reduce the risk of failure. One attempt to curb the random nature of bore-hole drilling, which of itself may not be a bad thing,7 was to introduce subsidies on condition that the applicant employs a private geologist to make a hydrological plan which could be taken to the Service of Land Improvement (YEB). This option was not often pursued, primarily because of reservations about dealing with the public administration and the time that the process could take. It did however, introduce the ‘expert’ to the process and geologists along with private drillers were often employed by groups of farmers to identify and drill for water. Previously, many geologists were employed by the agricultural service and some drillers were state employees. By 1993 the majority of drillers and geologists were self employed and would work either on a time related contractual basis, or for a higher rate, they would require payment only if reliable water was found.8 Alongside the search for more reliable sources of groundwater has been the introduction, and expansion of, infrastructure for transporting Anavalos spring water from Kiveri in the south west of the case study area to the central plain.

4

Some farmers also filter saline water. A general estimate for the cost of a new bore-hole is around 5,000,000 drachmas. 6 T.O.E.B are the local arm of the Service for Land Improvement (YEB), although they only have responsibility for irrigation and water use. 7 See Allen and McGlade, (1987), re. Stochastic and Cartesian approaches to fishing. 5

122


THE EXPLOITATION OF SPRING WATER: ANAVALOS In the early 1960’s the central Argolid Plain suffered progressively from salination which had occurred primarily in response to sea water intrusion as ground water levels became lower due to excessive pumping for irrigation. Historical records show the existence of freshwater springs (Anavalos or Dini) with water originating from the Arkadia mountains and emerging into the sea at Kiveri (Genesio) on the Argolic Gulf to the south west of the region. In the early 1960’s, in response to the increased salinity, it was proposed that the water from these springs could be captured within a sea wall or dam and transported by canal to the plain, a distance of approximately twenty kilometres. The initial plan was rejected by the Ministry of Agriculture in 1966 because it was feared that dam would not prevent sea water intrusion and that the water transported for irrigation would be too saline. In 1967 with the arrival of the Junta this decision was reversed and construction started on the dam under the supervision of German engineers. The initial plan put forward by the engineers stated that the water would contain a maximum of 300 parts per million (ppm) of sodium chloride (Papaioannou, 1980). 165ppm is the accepted limit for human consumption and 500 ppm causes considerable distress to citrus plants (Wallace, 1976). Salt content in the Anavalos water has reached 600 ppm in periods of drought, although, this is substantially reduced when there is increased precipitation. The springs have an output of 12m3/second providing a maximum quantity of water transportable to the valley for irrigation purposes of 150×106m3/year (AUA, 1994). This equates to approximately 90,000 metric tons of NaCl or 3,500kg/ha/year which threatens to make much of the soil unsuitable for agriculture if there is insufficient leaching from rainfall. The salination of groundwater described in chapter eleven posed a considerable problem for agriculture in the early 1990’s. This has, however, reduced since 1994 with the increase in precipitation and the expansion of aquifer recharge using fresh springwater from Kefalari (see below). There remains considerable optimism about the role of the scheme for the transportation of spring water to areas suffering from salinated water, or from water shortage. This optimism is reinforced by the move towards ground water recharge from the freshwater springs of Lerni and Kefalari. Construction is currently underway to extend the Anavalos canal to Iria in the east of the area for replenishment as opposed to irrigation purposes (Figure 7–6). Without questioning the technical argument in favour of the Anavalos canal it is possible to identify a paradox between the role of the project as a psychological prop, which has encouraged local hope, and a reluctance to use the scheme when independent options are available, even if these are short term. This attitude is conveyed by a farmer from Kefalari (zone 2) “I irrigated with Anavalos for the first two years after I bought the farm. Then I opened a bore-hole which is easier to irrigate with and allows me to irrigate whenever I want. There is a difference in quality as well, the water from the bore-hole is better than that of Anavalos. The water from Anavalos is good, it may be a little salinated but it does not harm the trees. Many people irrigate with Anavalos, the ones who do not have their own bore holes.” The final statement from this quotation is interesting in the light of recent (1997) aquifer replenishement, partly from increased rainfall and partly from artificial recharge using the Anavalos

8

I.e. a continuous flow for one year.


123

Figure 7–6 Anavalos pump house and dam.

water. The implication of this statement is that whenever possible farmers will use their own boreholes rather than the Anavalos infrastructure for irrigation water. This means that the system is difficult to monitor and control with the likelihood that excessive water use will result in a revival of salination and water depletion. There is also a scepticism attached to the scheme which has taken thirty years to complete and is perceived as having been a political instrument for every government since it began. Political expediency is seen to have determined a course for the canal which has left the areas around the periphery of the valley without water from Anavalos. “We have been told that Anavalos will come to our village since 1981 but nothing has happened. Even if we want to bring it here ourselves they (the government) will not let us. Monisteraki asked to bring it there at their own expense but they said no. It is used politically by PASOK and ND, everybody tells us I will bring water”. (Interview with field guard, Koutsopodi)

Other factors that either restrict the use of Anavalos or reinforce the use of ground water sources are that Anavalos water often cannot be used for protection against frost because the water is too dense to pass through the fine sprinkler heads. There are also incremental prices depending upon the number of irrigations. If more than twenty five irrigations are undertaken then the price can rise by 25%. This means that there is a differential price for irrigating different crops, i.e. in Kefalari oranges cost 3,000 drachmas per stremma for eight irrigations, whereas vegetables cost 3,750 drachmas for the same number of irrigations because the total number per annum will exceed twenty five. As has already been stated the water from Anavalos is also too dense to be used in

124


Figure 7–7 Section of the Anavalos distribution canal.

many of the installed sprinkler systems. This means that less economic application procedures may be undertaken i.e. free flow. It may also explain the predominance of sprinkler systems in the peripheral zones where Anavalos is not always available. The combination of these factors explain the continual use of other water sources and the perpetuation of fruit tree production as the basic crop where Anavalos is the main source of irrigated water, in zones five and six. Anavalos, therefore, is currently reinforcing the monocropping system in the Argolid plain and where there is more crop variety other sources must be used. If this alternative is salinated then the crops must be more tolerant of salt i.e. artichoke and lettuce. If the water is of good quality then the cost of Anavalos may well prevent its uptake.9 Those areas that have severe problems of supply (i.e. Mykenes) currently do not have access to the Anavalos system.


125

Artificial Recharge Figure 7–8 shows those bore-holes that are currently being used for artificial recharge, it also indicates the extension to the Anavalos (Kyveri) canal for aquifer recharge around Aria. Recharge techniques using fresh spring water from Kefalari and Lerni were suggested in the mid 1970’s (Wallace, 1976) however experiments were only commenced by the Agricultural University of Athens for the Ministry of Agriculture in 1992. Farmers were often reluctant to commit their bore holes, even those that were salinated, to the experimental process. As a result it took a considerable time to gain their trust and extend the programme into the valley. However, the flooding which occured in 1996–97 has again led to a questioning by farmers in the valley of the recharge process (see below). Sprinkler Systems Sprinklers have been used both against frost and for normal irrigation in preference to free flow. Table 7–2 shows the distribution of the two types of irrigation from the interview sample spread across the Argolid Valley. Table 7–2 Types of irrigation application by zone.

Zone six which suffers from salt and frost has a far higher proportion of free flow than would be expected for several reasons. The area is the main recipient of Anavalos water which often cannot be used with sprinklers because of time restrictions upon use which may not coincide with low temperatures. More important was the high salt content in the aquifers which meant that the water was unsuitable for irrigation. Paradoxically this also led to the use of sprinklers for applying salinated ground water from bore holes to protect against frost and may help explain why over 50% of respondents in this zone used sprinklers. It is also noticeable that zone five has only one respondent who irrigates using free flow only, although seven use both application techniques. Again this allows for Anavalos water to be used while retaining the option to spray against frost. Those areas that suffer less from frost and do not have access to Anavalos, or do not need to use it because their ground water is less saline than that of zones five and six, will be more concerned with water conservation. The poor soils and slope of the peripheral zones mean that more water is required for irrigation. For example if sprinklers are used in the valley 100–150 tonnes per stremma will be used to irrigate oranges; in the foothills this figure will increase to between 300 and 400 tonnes per stremma. If free flow irrigation was adopted, however, this figure could again be doubled. Wherever possible, sprinkler systems are installed at the time of planting a new crop. It is interesting to note that systems were installed in the peripheral villages a year or so before those of the central plain (Table 7–3). This decision may have been 9

The cost varies but 2,500 drachmas/stremma/year and 800 drachmas per hour is realistic.

126


Figure 7–8 Map of the artificial recharge network.


127

Figure 7–9 Sprinkler systems are used for irrigation and against frost. Table 7–3 Mean date for the installation of sprinkler systems by zone.

swayed by the opportunity to irrigate more economically, in terms of the amount of water applied, with sprinklers and the availability of subsidies for the installation of systems after entry to the European Community in 1981. The peripheral villages tended to suffer more from depleted water stocks than declining quality. The opportunity to reduce the amount of water used was therefore more important to farmers in these areas than it was to those from the valley who were more concerned with the quality of their irrigation water.

128


FLOOD CONTROL An ironic, but important problem in an area that has experienced problems of water depletion is the recurrent one of flooding, either in the form of periodic flash floods following very heavy and intense precipitation, or as a result of prolonged periods of rain. Respondents felt that the former occured every twenty to twenty five years whereas the latter events were more frequent. Representatives of the victims of recent floods in Argos (1996–97) claimed that they had experienced similar problems eight times since 1967. The heavy rainfall between the autumn of 1996 and early 1997 resulted in considerable flooding in the reclaimed wetland areas of Nea Kios and further inland to the Southern part of Dalamanara on the main Argos-Nafplio road. This in turn led to the loss of much of the orange crop in these areas and in some cases to the trees themselves. The floods led to a number of accusations and counter-accusations between the farmers, residents of Argos, the political elite and the Universities. The former were seeking compensation for damage to their orchards and felt that the floods were the direct result of the aquifer recharge work which had in effect restricted the undergound storage capacity and forced the excess water to the surface. The university in turn felt that much of the problem was due to the filling in of run-off ditches with soil by the farmers. In the urban area close to the Xerias river the accusations focused upon the sale of land parcels in the dry river bed by politicians and the subsequent planting of citrus trees in these parcels. The response to the floods has highlighted one other factor that is important to the Argolid study and is seen as key to environmental management, this refers to the lack of an integrative and coordinated approach and to the perceived ‘ownership’ of courses of action and the information relating to them. There are clubs of professors with no understanding between them, half of them are against the other half. Politicians are mixed up in the game and a mess is created. (Interview with an agronomist from the Service of Agriculture, 1993). In the context of flood control specifically and water management more generally, a number of related points can be made which refer to the competing technical ‘solutions’. There would appear to be competing academic responses between hydrologists who favour the use of the aquifers as reservoirs and advocate replenishment and those who argue for the surface management of water through the building of dams and reservoirs. On a political level there is also a perception among the local farming population of conflict between Regional government support for diverting the Xerias river and that of the Urban authorities who favour the reservoir and dam option. FROST, TECHNOLOGY AND WATER USE As has already been seen, sprinklers are often employed to spray against frost.10 The alternative to the use of sprinkler systems to protect against frost is the installation of air mixers. The spatial and economic implications of these will be discussed in greater detail below, after an introduction to the issues raised by the problem of frost in the area. Because of its high salt content Anavalos water is often too dense to pass through the sprinkler nozzles farmers may use water obtained from bore holes to spray against frost, even though the water is salty. This is an important element of the production, and hydrological, systems because little information is available about the extent to which sprinklers are used for this purpose. The use of


129

ground water in winter prior to 1994 was underestimated and hindered the recharge programmes which were planned around minimal water abstraction and maximum rainfall retention. With the increased rainfall of 1994–97 less concern was expressed about the need to conserve water as a central feature of successful recharge. The importance of this in the earlier period was evident when there was low rainfall combined with a long period of recurrent frost (i.e. 1991). In some instances it was possible to as much water for frost protection as for summer irrigation. This is well documented, as are the behavioural implications arising from it, in the following extract from an agronomist employed by the Department of Land Improvement (YEB). Again the changed situation with more rainfall must be borne in mind, however, the implications for local decisions concerning the management of natural resources remain significant. “Let us talk about replenishment, this must take place during periods of no irrigation. There are no such periods anymore. After November we start to use sprinklers for protection against frost. To make our (farmers) life easier we start the sprinklers at 0°C or even 1 or 2°C, so that we do not have to get out of bed or so that we can play cards in the cafe without being disturbed. Some sprinklers may well work unnecessarily for the 100–150 winter nights when the temperature drops to 1°C. So we have an irrigation period from the 1st January to the 31st of December, you tell me when the ‘dead’ period is when we are going to undertake replenishment. If we replenish during the winter time then we are going to re pump this water the next day for protection against frost. Thankfully we save some water by irrigating with sprinklers instead of free flow, but these other problems come out of this”. A similar point was made in a seminar on the Argolid water problems run by the Association of Agronomists in 1992. This also raised an important point that is often underestimated when adopting a technical perspective to assess technological potential. Namely, what can be done if the people using the technology do not do so in a manner which is ‘appropriate’ to the technical perspective but understandable in its cultural context? Floor question: “Farmers are using sprinklers in winter for protection against frost, however, the problem of frost only arises occasionally. The rest of the time the water is used needlessly”. Response: “The use of sprinklers does protect the plant against frost, however, if it is being used at times when there is no frost then I do not know what to say”. (Argolid Association of Agronomists, 1992) The majority of data relating to water consumption do not incorporate figures for water use against frost. There is a body of opinion within the farming community that the frost problem is in fact self perpetuating. The more leaf cover in an area, the higher the level of humidity, and therefore, the greater the risk of frost. Some crops that were grown extensively throughout the area i.e. lemons, can no longer survive because of the frost levels. One theory, therefore, is that the increase in frost is the result of micro-climatic changes resulting from extensive irrigated fruit cultivation. A second proposition is that frost has increased with the removal of the stagnant water which lay alongside the coast between Nafplio and Nea Kios. This water was a breeding ground for 10

Sprinkler systems use approximately 150–200 litres per hour, in large drops (Margaritas) to irrigate and 70 litres per hour, in a spray to protect against frost.

130


mosquitoes and thereby malaria, it also maintained the local climate at a slightly higher temperature than is currently the case and as such protected against the danger of heavy frost. It is possible that both of these hypotheses have an element of truth and that the situation is similar to that which has been recorded in Arta (north-west Greece) where the removal of wetlands and a dramatic increase is citrus production have coincided with an increase in the level of frost. A third proposition was put forward by Wallace (1976) who supported the view that the increased frost was due to the expansion of citrus production. The Spatial Distribution of Air Mixers and Sprinkler Systems Used Against Frost Table 7–4 shows the distribution of hours spent using sprinkler systems to protect crops against frost. It appears contradictory that zone six which, with the exception of zone five, suffers most acutely from frost has one of the lowest mean figures. This is explained firstly, by the extensive use of air mixers to protect against frost in zones five and six, whereas the other areas that encounter the problem have far fewer, if any, mixers. Zone five has both a high use of water against frost and high mean figure for air mixers. The difference between these two zones rests primarily with much of zone six being close to the sea with a warmer local climate and less frost, whereas the rest of the zone, however, is well protected by air mixers. Zone five does not have a sea border and even with a 30% to 40% coverage by air mixers still requires additional protection from sprays. Zones two and three appear to have a higher mean figure for water used against frost than might have been expected. This can be explained by a sampling bias of farmers who work land in the south and east of Koutsopodi and east of Argos where frost is a problem, and where there are fewer air mixers than required. Table 7–4 Mean number of hours in which spray is operating against frost, and mean figure for air mixers per farmer for each zone.

The first air mixer to be installed in Greece was in Pyrgella in the Argolid Valley in 1967. This was brought to the area by the Ministry of Agriculture from Kotonia (Italy) for experimental purposes and remained in the port of Piraeus for some time because no farmers were prepared to have it installed. The relevance of this point extends beyond the diffusion of technologies into all aspects of decision making, particularly crop choice. The options that are available to farmers must be considered alongside the social mechanisms that support their take up (i.e. the existence of a local risk taker or a farmer who is sympathetic to innovative ideas and technologies, and subsequently the importance of peer group networks). The choice of Pyrgella for the installation was also interesting because from the farmer’s own admission other areas in Greece such as Arta in the North-west and Mystras in the Peloponnese had worse frost problems. As indeed, did other areas of the Argolid valley. The farmer who first installed the air mixers had a history of contacts with the Ministry of Agriculture and had made it clear that he was willing to participate in agricultural experiments. This


131

Figure 7–10 Air mixer in the Argolid Valley.

innovative sympathy was (is) not common within the farming community and when it did exist it is possible that experiments could have followed the path of least resistance rather than being directed at that which was the most appropriate in terms of experimental design. When the first air mixers were introduced on an experimental basis the operating costs were paid by the Organisation of Farm Insurance (OGA). This was subsequently terminated and the petrol driven mixers cost more to run than the farmers were prepared to pay (i.e. in 1992 one air mixer cost approximately 30,000 drachmas per year for petrol or electric in a year with frost, and 120,000 drachmas for oil and maintenance costs). This resulted in the mixers not being used and the adoption, and resiting, of many by the co-operatives. In 1992 one hundred air mixers were obtained by the state from the Sparta area (near Mystras) and in 1997 a further batch were made available. In

132


Figure 7–11 Elliptical coverage by an air mixer.

both cases the distribution of these mixers has been the source of conflict. There was some concern that the ownership of the mixers by the co-operative and their distribution would at best be inappropriate or at worst according to political sympathy rather than the extent of frost in an area. Over the past three years frost has become an increasing problem in the peripheral zones which have not had access to the ‘subsidised’ mixers. Consequently farmers in these areas have had to collectively purchase mixers if they require this form of frost protection. Because air mixers cost 6,000,000 drachmas or £20,000 (1993 prices) the decision to purchase is invariably a collective one. It is also a difficult one to negotiate between prospective partners. This is because the mixer provides an elliptical cover across forty stremmata, depending upon the prevailing wind conditions (Figure 7–11). This coverage obviously does not coincide with specific units of land and as such will inevitably result in ‘free riders’ or a disproportionate advantage for one partner over another that may need compensating for. To be of real benefit air mixers need to be installed in an integrated group rather than in isolation and this emphasises the need for some form of collaborative arrangement. There are a number of financial benefits from the ownership of air mixers, over and above the protection of the crop and tree from frost. The co-operatives will often have a different contract with farmers possessing air mixers. This allows the producer to sell their oranges elsewhere but charges a five drachma per kilo fee to retain the option to sell through the co-operative. The growing season is also extended because the crop does not have to be picked at the peak season to avoid early frosts, this facility is accompanied by higher supported prices from the co-operatives. An exception to this rule occurred in 1997 when the supported market declined and many farmers held onto their crop in the hope of selling at a later date for a better price. These competed with the late produce from farms with air mixers who traditionally had received higher prices. The competition meant that the price for late oranges dropped considerably and many producers failed to sell all of their produce. The financial impact of this was obviously felt more acutely by those who had met the running costs (technology, maintenance and electricity) required to protect their crops against frost.


133

CONCLUSIONS AND DISCUSSION This chapter has examined the role played by technology in the emergence of agricultural production over the past forty to fifty years. It has attempted to do this by describing the spatial and temporal development of those technologies and placing them in the context of the system as it is perceived by the farming and scientific communities. If we look at the chapter in a wider context a number of points can be highlighted about the role of agricultural technology in the Argolid: 1. Technology has enabled the expansion of irrigated agriculture and is also seen as the predominant solution to the problems of natural resource degradation that have arisen from this practice. 2. The spatial distribution of technology has led to variations in the pace of change which have resulted in an inequitable distribution of costs and benefits within the valley. 3. The restrictions relating to certain technologies have reinforced the over use of water (i.e. the spatial limit for establishing bore-holes may have led to more, rather than less, being drilled). Similarly the support provided for irrigation technology through reduced electricity prices also encourages greater water use. 4. There is inadequate data about the extent of water use, particularly in terms of frost protection. 5. Too little account is paid to the social influences relating to decisions about the adoption, and use, of technologies. This is particularly relevant to the adoption of individual, collaborative or ‘state’ funded technologies. While it is not appropriate to argue for one technical solution over another it does appear to be the case that the period of low rainfall in the late eighties and early nineties had focused attention on water depletion and quality to the exclusion of problems associated with flooding. Current concern however relates to the control of flood water to the exclusion of depletion and water quality issues. This combination of technological fixes and short term perspectives is one that tends to prevail to the exclusion of a more flexible approach based upon integrated community, political and scientific interests. A further question arises out of this discussion on the technological underpinning of agriculture and the production history outlined in the previous chapter. Technology has widened the spatial potential for irrigated agriculture alongside increasing uncertainty, firstly in terms of degradation processes and secondly through the reduction of the artificial market place. One response to this has been identified as a move away from citrus production towards alternative tree crops such as irrigated olives. A similar argument can be put forward about the use of information technology in support of an apparently contradictory trajectory, namely that which facilitates more ecological approaches to farming. Of course access to information technology is not restricted to the ecological option but it does raise the important point that the role of technology is both hardware (pumps, canals etc.) and software (information, management techniques etc), and that the utilisation and adoption of either of these is dependent upon the social and cultural context within which farmers operate. The next two chapters will consider this alongside the structural characteristics of agriculture in general and the Argolid Valley in particular.

134

8. STRUCTURAL WEAKNESSES AND THE ARGOLID? Mark Lemon and Nenia Blatsou

It was noted in the introduction to this study that Mediterranean agricultural systems often do not conform to Northern European models of production and economic efficiency. Indeed it was suggested that when these models were followed then considerable pressure was placed upon natural resources (Ruiz, 1988) and the local socio-economic structures that were reliant upon them. This chapter will look in more detail at the structural characteristics of agriculture in Greece as a whole, and the Argolid in particular, and in so doing will supplement the bio-physical and technical components of agricultural decisions with a socio-economic and cultural perspective. These will then be considered, in the following chapter, as the constituent attributes of a crop choice framework which represents the variation in decision making across the study area. It has already been argued that the location of the Argolid Valley has helped it avoid the marginalisation that has affected other rural areas in Greece. This has enabled the area to establish and maintain contact with local and European markets. The area is also close enough to urban centres for alternative sources of work and income generation. Although this will be seen to be of particular importance in the classification of a range farming types, however, for the time being it is sufficient to observe that while land-use in the Argolid Valley is based upon agricultural production the area itself is inseparable from a range of urban based networks. The Argolid will be seen to have developed an economic base that is dominated by agriculture while the contribution of agriculture to the Greek GDP and the investment to support this, has declined markedly in importance over the last twenty five years (Tables 8–1 and 8–2). This is particularly evident when it is compared with the other central areas of the Greek economy in which the manufacturing sector has remained relatively stable while the service industries and public administration have increased steadily. The declining share of agriculture in the overall economy has been accompanied by a steady movement from rural areas and from the agricultural sector, to urban areas and more sedentary work (Table 8–3). This movement has similar origins to those which generated the mass exodus Table 8–1 Share of GDP by industrial sector.

136


Table 8–2 Share of total investment by industrial sector.

Source: OECD 1991/92. Traded services refer to transport, communications and other services. Non traded services are dwellings and public administration. Table 8–3 Trends in the rural population of Greece, 1921–1981 (millions).

Source: National Statistical Service of Greece.

overseas in the 1950’s, 60’s and 70’s.1 They lay partially with the difficulty in generating sufficient earnings from agriculture and in part from the low status that is often attributed to farm work (De Waal, 1991). The population of the Argolid Valley, however, has remained stable over the last twenty years with 45,000 inhabitants in 1971 and 48,500 in 1991. At the same time the area has retained a high percentage of its work force in agriculture with around 50% employed in the industry compared with under 30% for the country as a whole (Figure 8–1). This trend must be seen in the context of the high levels of multiple job holding or pluriactivity that are evident in Greek agriculture and which will be discussed below. Therefore the area has utilised its geographical position and restructured its agricultural production while retaining many of the characteristics that elsewhere have been seen to exemplify inefficient farming and thereby to explain the inability of Greek agriculture to compete in a wider market. Whereas elsewhere this process has often resulted in rural depopulation, in the Argolid it has led to an increasingly vulnerable natural resource base. A broad consensus is evident about the structural weaknesses that are inherent in Greek agriculture and have prevented its efficient integration into a wider market system, particularly since accession to the European Community in 1981 (Polopolus, 1989; Spanopoulou, 1990; Damianos, 1991; Demoussis and Sarris, 1988). Greek agriculture is at a junction between two stereotypical models of agricultural production. The first is based upon a ‘traditional’ model of self sufficient farming with minimum inputs, small, dispersed farming units and surplus production sold through local markets. More recently, and particularly since accession to the European Union, a new model

1

During the period 1956–1975–2,163,250 people migrated from Greece (Spanopoulou, 1990).

STRUCTURAL WEAKNESSES AND THE ARGOLID?

137

Figure 8–1 Argolid occupational structure.

has emerged based upon economic efficiency, high inputs, investment in technology rather than labour, larger units and less diverse production for an external market place. The ability of one model, the ‘traditional’, to respond to change is inevitably restricted if the objectives of the other are seen to prevail. As a result the emergent model must be considered in the context of the diverse socialeconomic needs of the farming community and the structures in place to meet these. As has already been seen in chapter two it is necessary to obtain information about the agendas of that community and to consider these alongside the range of structural options (Lemon and Park, 1994). This will be considered in terms of the components of farming decisions in the next chapter. For the present, however, we will concentrate upon the structural characteristics of agriculture in the Argolid and Greece as a whole. These will provide the background for understanding how land-use has changed in the region and subsequently what options appear (in)appropriate because of this structural configuration. A range of ‘structural weaknesses’ have, therefore, been identified as constraining the ability of Greek agriculture to conform to the requirements of a ‘modern’ and increasingly standardised industry. These include: • • • • • •

small fragmented land holdings multiple job holding an ageing, poorly educated farming population insufficient rural infrastructure an inefficient bureaucracy inefficient marketing organisations

Policies have been introduced to address some of these structural issues2 although a dissenting view must be recorded which questions the relevance of this industrial model (Fordism) for local, small scale farmers.

138


Fordism wove a powerful web which sought to ‘teach’ peasants to regard Fordist technology and consumption as a mark of superiority. This explains why it succeeded unchallenged in bombarding an area of the Greek economy that had only limited scope for improved productivity under costly inputs designed for scale economies. And so, when it raised the social cost of agricultural activity and set an aim foreign to it, the aim of standardised output, which peasant agriculture could not effectively achieve, the blame fell on the ‘anachronistic’ structure of the peasant economy (Tsapatsaris, 1986). It will be argued, in terms of both structural characteristics and farmer types that agricultural production in the Argolid is not consistent with either the traditional or Fordist models. What is more important, however, is to question how the production system has evolved in such a way that it is particularly vulnerable to a degrading natural resource and to changes in an artificial market. FARM SIZE AND LAND OWNERSHIP Greek agriculture is largely built around small farms (see Tables 8–4 and 8–5), often comprising of a number of dispersed units. Unlike many northern European countries where farm size is primarily determined by productive capacity, in Greece the political and social role attached to land ownership has been of prime importance (Lianos and Parliarou, 1986). Until early this century land was passed onto the eldest son and this retained ownership, continuity and farm size. However, more recently property has been divided between offspring and where the inheritors do not wish to pursue farming then their share can be purchased by the others. Therefore, the small size of farms Table 8–4 Farm size in the European Union (1979).

Sources: Eurostat, 1985, Vol. 111 Table 7.3; 1986, Table 7.3; Greek Statistical Service.

can been linked to the partible inheritance laws in Greece, which have encouraged a more equal distribution of land amongst heirs; and to previous land reforms which have redistributed state, feudal and church lands amongst landless peasants and refugees.3 It was anticipated that such land

2

See Council Regulation (EEC) No. 797/85 on improving the efficiency of agricultural structures.


139

Table 8–5 Farm size (stremma) and parcels of land per farm.

Source: *National Statistical Service of Greece, **Survey of 203 farmers across the Argolid Valley.

redistribution would not only secure a more stable social system but that it would raise productivity by concentrating upon small units with self motivated farmers rather than large and inefficient estates. Forbes (1976) argues that land fragmentation ensures the equal division of property between heirs and is considered to provide some protection against natural hazards. For example a hail storm may affect one parcel but not another even in relatively close proximity. Similarly the impact of frost can be very localised. Paradoxically, recent policy has tried to reverse this process with the argument that efficient agriculture would benefit from economies of scale that are not open to the small farmer. Attempts to pursue this through land restructuring programmes (Anathasmos) have invariably failed and will be discussed below. When a landowner dies or a piece of land is sold then the legal procedures are carried out by state certified property lawyers. If no will has been made then a statement of inheritance is made by all the claimants and when agreement is reached a contract is drawn up by the lawyers. All land deals, including inheritance, are subject to tax which is payable at the nearest tax office in Nafplio or Argos. The tax charged on a property deal varies with the value of the property or land, its location and whether or not the purchaser/inheritor is already working on the land in question, or elsewhere, as a farmer. When this is the case no charge is levied for the first twenty stremmata although the standard rate is applicable after this. Until the late eighties it was possible for forty stremmata to be passed onto each child on the condition that the land would not be sold for ten years. If a sale did occur within that time tax was payable before the sale. The rate of tax for purchasers from outside agriculture is between eight and ten percent. Some attempt has, therefore, been made to retain ownership in the hands of farmers through land transfer procedures. What this has failed to account for are the social changes that have accompanied the recent agricultural transformation in the Argolid. Central to this has been the emergence of a significant group of part-time farmers for whom farming provides a secondary source of income and who are often restricted from working the land in a sustainable manner because of time constraints, insufficient information and knowledge and an unwillingness to make personal or family commitments to farm work. The low status attached to farming and the commitment to other employment outside of agriculture by this group prevents the adoption of more labour intensive crops.

3

For example Law 2158/1952, see also Forbes (1976).

140


Figure 8–2 Example of three generations of Argolid farmer.

Therefore, since the 1940’s four interlinking changes have occured relating to farm size in the Argolid Valley. • • • •

There are less large land-owners There are more farms The farms are smaller The farms have been divided between families

Each of these trends have to be seen in the context of changes in land-use and the inheritance laws and are represented in the example in Figure 8–2 which portrays the changes of a family farm in the Argolid over three generations. This provides an insight not only into changing farm size and the increased number of units, but also into changing land-use with the arrival of less labour intensive but highly profitable fruit tree cultivation. This has accompanied the increased availability and use, of irrigation water and places less emphasis upon the land owner to be actively employed in production. Villages in the plain (zones two, five and six) have therefore been able to move from rain fed extensive agriculture to intensive fruit production with the result that individual units of land have provided maximum returns for less labour. This in turn has reinforced the position of the part-time farmer. Table 8–5 shows the Argolid to have fewer very small farms (less than one hectare) than the Greek average and a greater number of farms between five and twenty hectares. This is slightly misleading because many of the bigger farms have large areas of poor quality and uncultivated grazing land in the foothills. Table 8–6 make this clear with the peripheral zones in the foothills and mountains (one, three, four and seven), having bigger farms as well as larger individual parcels of these areas are more diverse in their cropping and land.4 It has already been seen that farms in have large areas of land that are not cultivated or used for less intensive dry farming and grazing. In 1993 the income generated from a stremma of land was higher in the central zones. The region was not, therefore, compliant with theories espousing the benefits of economies of scale. What is more important are the physical characteristics of the land, the availability of water and the time that is spent by farmers on agricultural work. It is difficult to analyse the dispersal of farms because this often occurs over a wide area.

4

With the exception of zone one where the mean parcel size is eight stremmata.


141

Table 8–6 Farm and Parcel Size (stremma=1/10 ha.) by zone.

Source: interview data.

Other Forms of Land Ownership In the early years after the removal of the Turks in 1821 much of the land was redistributed from the state and the church to poor farmers and those who had worked for the revolution (Thompson, 1963; Kofiniotou, 1892). Between 1821 and 1911 approximately three million stremmata of church land was redistributed and a further seventeen million was transferred to 300,000 recipients from other sources between 1917 and 1936 (Petsalis, 1948; Sideris 1934). Among the beneficiaries were 150,000 refugee families from Asia Minor. In the Argolid these were settled predominantly in Nea Kios on the coast between Nafplio and Argos. Also under Law 2158/52 a further 5,500,000 stremmata has been removed from private estates and churches across the country since the Second World War. In the Argolid the church still owns a small percentage of the land in the valley and within this there remain some sizeable farms e.g. 300 stremmata in Skafidaki is owned by a church from Arkadia. There is also an amount of common land (woodland) that belongs to the state and cannot be sold. If it is close to a community the common land may be rented by that community for the grazing of sheep and goats. Some areas have environmental restrictions (i.e. of natural habitats and of course archaeological sites) and it is not unheard of for fires to be started in an attempt to destroy the habitat and overcome the building and cultivation constraints accompanying them. The amount of rented land in the area is relatively small, around one thousand stremmata. This is primarily because of legislation which effectively extends a one year contract by several years should the tenant wish to do so. If a contract is made for one year then, under Greek law, the tenant can work the land for four years and stay for two additional years. This results in an informal rented sector which has been used predominantly for the growing of tobacco and vegetables, particularly in those areas where a crop (e.g. apricots) has been uprooted in response to a restructuring programme and there are vacant parcels of land. An approximate charge for one stremma of land for tobacco, over a five month growing period and with a water supply, was 60,000 drachmas in 1993 and currently ranges between forty and sixty thousand drachmas depending upon the availablity of water. Elsewhere renting will generally occur where the land requires a lot of work, i.e. in the coastal area between Argos and Nea Kios, particularly where there is water available for irrigation.

142


Attempts at Land Redistribution—Consolidation (Anathasmos) Land ownership in the Argolid has changed considerably over the past fifty years in ways that are contrary to general European trends. Other Mediterranean countries (i.e. Spain and Italy) also have small farming units, but often with established collective agricultural structures (around one hundred stremmata, or ten hectares). These have enabled them to respond to agricultural legislation that is invariably aimed at larger scale farming units.5 In the Argolid this has not been the case and collective organisations have mainly been established for marketing purposes (co-operatives) and for capital investment i.e. in technology. There have been some Greek programmes (Anathasmos) to join disparate units of land into single larger units. These have occurred mainly in the mountain areas of the Argolid but with little success because the land varies considerably in quality, location and accessibility. This meant that people could not agree upon a restructured format for their parcels of land. In the valley effective restructuring is unlikely because of the variation in investment, both of time capital, in the land. It also depends upon the age and productivity of the trees, (as a general rule the older the tree the greater the production), the quality of soil, water quality and micro-climate i.e. frost and the time of production,6 and the availability of various technologies e.g. air mixers, sprinkler systems. In the valley the price of the land is also largely determined by access to a road, village or the sea. Although there are many main roads there are also many pieces of inaccessible land and the need to install infrastructure reduces the value accordingly. Farms in the central plain of the Argolid decreased in size until the early 1990’s. This was partly due to the increase in intensive fruit production which can be undertaken by part time farmers and the influence of the inheritance laws. However, if the system comes under even greater strain and it is necessary to spend more time or invest even more heavily in agriculture then larger farms may result. Land Values and Land Use Planning Land values in the Argolid Valley vary and fluctuate considerably. For example, sea front sites, for tourism, could be sold in 1993 for approximately five million drachmas per stremma whereas previously they were of low value because of their poor agricultural utility i.e. as grazing land. Areas with mature fruit trees and water also had a high value particularly if the land had planning permission. Sites near the Argos-Nafplio road with permission to build could fetch three to five million drachmas per stremma and those without permission, two to three million. The surrounding hills with less water and few trees may have only fetched half a million drachmas for one stremma of land. As has already been seen, land values are also influenced by the technologies that are in place to improve or protect agricultural production i.e. air mixers, sprinkler systems, bore-holes. For planning permission to be granted on land that is not close to the city, the total area in which the proposed development is to be situated must exceed four stremmata. Nowadays people only infrequently take the risk of not obtaining permission. This was more common before the electric and water companies began to refuse to connect supplies to properties without planning permission, thereby operating a similar control to the one which exists for restricting the opening of new bore-

5 6

Interview with director of Argolid Union of Young Farmers. Early and late crops can bring in better prices by extending the duration of the market.


143

holes. The state also has the option of demanding financial compensation or destroying buildings if planning permission is not obtained, however the risk is most likely to be taken in tourist areas, such as Tolon, where considerable rewards can be accrued. The total number of tourist properties built upon farm land with or without planning permission is, however, relatively small, with no more than one hundred in the area situated around tourist attractions of Tolon, Assini and Mykenes. In 1994 it could be argued that the slow movement towards agrotourism was due partly to a reluctance to change from what was currently profitable (citrus production) and into something which is seen as culturally distant (service industry). There was little perceived need, or advice and support for this form of tourism. In 1997 the changes in agriculture resulting from uncertainty about price support for citrus production has meant that the existence of a secure income is no longer a viable reason for not investing in tourism. The changes that have taken place however have been in the type of agriculture carried out in the valley (i.e. increase in irrigated olives) and the relative increase in peripheral land values when compared with those of the central plain which have decreased considerably. The extent of this change is emphasised by the value placed on a unit of land by the tax office (Eforia) under the Ministry of Economics. The price set by the tax office, for land tax purposes, had previously been lower than that which was achievable on the open market. Since 1995 the situation has reversed with the market setting considerably lower prices than the tax office (i.e. Interviewees in Poulakida stated that the tax office figure was 1.5 million drachmas per stremma for land that would only fetch 500,000 drachmas). This directly affects those who are paying high land taxes on parcels of land that have a far lower realisable price. The relatively high price of oranges sold to the open market, as opposed to those qualifying for price support, by farmers on the periphery and their greater propensity to farm in a more diverse manner has meant that the value of land around the plain has increased in value both in real terms and relative to the centre. MULTIPLE JOB HOLDING IN GREECE The size of agricultural units and the type of production is inseparable from the employment structure in place to support it. For example, the adoption of local markets to sell produce on the one hand, and pluriactivity on the other, both require the members of farming families to “rearrange their domestic roles in such a way that the family farm and farming is maintained”. (Pires, 1988). Multiple job holding is an important feature of agriculture around the Mediterranean and in Greece in particular (see Tables 8–7 and 8–8). However the role that it fills can differ fundamentally both between areas and between farming groupings in the same area. These differences depend primarily upon the proximity to urban, industrial and tourist activities, both as sources of alternative employment and markets and as influences on the quality of the natural resources (EfstratoglouTodoulou, 1988). These are important points for understanding the way that agricultural production has been structured in the Argolid and guard against any homogeneous interpretation of the role played by pluriactivity in the area. Pluriactive farming tends to occur on farms that are smaller than those which rely solely upon income from farming. In Greece 50–55% of those farms under two hectares in size are worked by

144


farmers with other sources of income (Todoulou, 1988). Table 8–7 shows a higher than expected population of multiple job holders in Northern Europe. However, if we adjust these figures to take account of the total area farmed by unit size then the picture changes (Table 8–8) and there is significantly more land farmed in small units by multiple job holders in Greece than in any other European Union country. Other general characteristics attributed to pluriactive farms are more educated farmers, less farm diversification, less labour intensive work with a seasonal work force (Todoulou, 1988; Sakellis, 1985). Each of these factors is evident among the citrus producing population of the Argolid which has a high level of multiple job holders, particularly from the professions, small business and property investment.7 Table 8–7 Percentage of farmers with other sources of income by size of farm (hectares).

Sources: Eurostat, 1985 vol. 111 table 7.3 and 7.9.

Table 8–8 Percentage of total area farmed by farmers with more than one source of income.

The changes in land distribution in the Argolid have resulted in the nature of the owner changing more than the ownership profile itself. Selling the land has been relatively uncommon because of the

7

47% of the sample of 203 farmers for this study had other sources of income.


145

Table 8–9 Activities undertaken by sample of 203 farmers in the Argolid (%).

historical attachment between land owners and their land which is often based upon status (Polopolus, 1989; DeWaal, 1991). The increase in input costs and the reduction in price support alongside the cost of addressing a degrading resource base, has however raised the profile of this option for many farmers. Paradoxically, in the Argolid, the same factors have also constrained the market for land. The problem in the central valley remains that of finding a buyer although the reason for this is less to do with deteriation in water resources and more to do with uncertainty around price support and the reduction in prices. The most significant structural change in the Argolid has arisen out of the increased production of citrus fruit and the low, and seasonal, labour requirements to support this. Small units of land with fruit trees have been capable of generating a good income particularly if this was not the only, or the main, income of the household. This distinction between mono-cropping and multi-cropping has become even more pronounced with the decrease in price support. In 1997 small mono-croppers, particularly of Merlin oranges were no longer able to make sufficient profit to justify their expenditure and they have therefore started to reduce the level of inputs used in production (fertilisers, pesticides). They are also more reliant upon other sources of income which in turn reduces their potential to farm in a more diverse and labour dependant manner. This recent trend has therefore dramatically restructured the relationship between the periphery and the main valley with the former now being seen as more profitable, and not simply as more adaptable, as was the case in the early 1990’s. Where the owner does not wish, or is unable, to work the land they may choose to rent or more likely to run the farm through contractors who are employed to undertake clearly defined tasks i.e. pruning, picking, ploughing (see Table 8–9). Such an arrangement would be more difficult with labour intensive cultivations. In contrast, the application of fertilisers and pesticides and irrigation is generally undertaken by the farmer. The figure for book-keeping is of particular interest in the context of legislation for the improved efficiency of farming structures which obliges farmers to keep accurate records of their activities.8 Part-time farming, where the land owner uses contractors for labour intensive or highly skilled work, has coincided with the ‘professionalisation’ of the owners into other occupations. The training to support this has often been paid for by the profits from agriculture which the farmers have invested in the education of their children, and in the properties in which they live while undergoing this training (see Sakellis, 1985).9 This coincides with other studies (De Waal, 1991) which relate land ownership to status in a way that is not always evident with farm work itself. Where citrus trees dominate the local agriculture the trees now resemble a production line and this has changed attitudes to farm work, particularly amongst the young. It has also changed the timetables

146


and routines by which people organise their day, with far less constraints than are apparent with other forms of farming. As a result there is less of a requirement for children to help on the farm and a corresponding decline in their knowledge about farming. A distinction has already been made between those farms that have a more diverse cropping and marketing structure supported by the extended family, particularly in zones one and three, and those that have less variation in cropping and whose children are less keen to enter farming. This latter group will tend to be more educated and to invest outside of agriculture. To extend this distinction to incorporate part time-full time differences is misleading because of the variation in part time farming activity. This is exemplified by a statement from a twenty five year old farmer from Pyrgella “We all know that most of the farmers do two or even three jobs because they cannot survive. This is not fair in my opinion. The real farmers have to share the subsidies with other people who have a farm but at the same time have an enterprise in the name of their wife or sister or do another full time job like driving a bus or a taxi”. A second model of part-time farmer therefore emerges, namely those who cannot generate sufficient income from agriculture, i.e. in zone four where poor soil and degraded water resources have meant that the farmers have had to look elsewhere for supplementary income. This model has to be updated to include those monocroppers in the central plain who have had a dramatic reduction in their income from citrus pruduction. The difference between these two groups is that those from the central plain are more likely to have established sources of supplementary, or even primary, income. In the former case multiple job holding is undertaken to support a farming vocation. However, the income generated from the secondary activity can become more important than that which is acquired through farming, particularly with less time spent on the latter. Therefore, the ‘farmer’ can become progressively more dependent upon the supplementary work and will find it increasingly difficult to return to full time farming (Damianos, 1991). This highlights a weakness in legislation which bases the qualification for benefits on the strength of the number of hours spent on, or the percentage of income generated by, farming activity.10 The problem is further compounded by policies that discourage farmers from seeking work elsewhere even when this is intended to support farming activity.11 Conversely it is possible for those who undertake intensive monocropping activity with a limited labour content to qualify for support. What has emerged, therefore is a distinction between those who are perceived as ‘real’ farmers and those who are not. This distinction will be developed further in the next chapter, for the present it is sufficient to suggest that it has less to do with the amount of income generated from agriculture than with a personal commitment to the lifestyle of farming. Regional policy aimed at achieving a more balanced growth between the regions and for maintaining the population of rural areas has been outlined through a series of Five Year Development Plans.12 These have advocated diversified employment opportunities for the farm

8 Reg. 797/85 (superseded by reg. 2328/91) on improving the efficiency of farming structures has the condition attached to all subsidies that adequate records are kept. 9 Farmers often purchase property in Athens which is used by their children while they undertake their studies. The children may continue to live in these if they remain in the city or the property may be rented out (see also De Waal, 1991).


147

population and this inevitably involves pluriactivity. These policies and programmes are less relevant to areas that are not seen to be suffering from depopulation or underemployment such as the Argolid. They also fail to acknowledge the internal variation in an area that collectively appears prosperous, particularly when the source of much of that prosperity, in the form of intensive agricultural production, is seen to be undermining the natural resource base of the area. In addition to this, many of the options that are proposed for the area i.e. new crops and agro-tourism, are not taken up because they are not appropriate to the social structures and the attitudes within the farming community. A reluctance to work outside of agriculture on the one hand and an inability to undertake more labour intensive farming on the other. AGE AND GENDER CHARACTERISTICS Reference has been made to the importance of extended family networks to support certain forms of agricultural production and conversely the limitations imposed upon some farmers by not having access to the labour of family members. The full time occupation of women in farming can be seen to be fairly consistent across the age range. This contrasts sharply with the male farmers who make up a far larger percentage of the work force over forty, a profile which mirrors the distribution within agriculture itself. “All the women here are farmers (Skafidaki). Only widows own the land officially, but I would say the women work more than the men here” (Village Secretary of Skafidaki, 1993). The figures relating to the female contribution are, therefore misleading because they do not account for the hidden labour that underpins many forms of agricultural enterprise, particularly where crops are sold at the local market and where produce is processed for domestic consumption. It also underplays the contribution of women to activities within a monocropping system, such as irrigation etc. The existence of the female labour force is therefore crucial for the adoption of more labour intensive options. Where farm work is not considered culturally appropriate by other family members these options are unlikely to be adopted. This indicates a clear distinction between the active role of women in the periphery as described in the example from Skafidaki and that pursued by them in the plain where their contribution is far less. The relatively high percentage of women employed in agricultural activity is therefore misleading in that it is an underestimate in the peripheral villages and an over-estimate in the centre of the area. The latter can be partly explained by the need to register wives as farmers when the husband has a primary occupation outside of agriculture. This retains the tax benefits and eligibility for price support and subsidies within the farm even though many of the women do very little in the way of farm work. This has a cultural basis that is expressed in the following statements by one of the few active women farmers in the central plain. 10 In Greece farmers are termed full time if they work more than 50% of their time (1,750 hours per annum) or earn more than 50% of their income from farming. 11 Only farmers whose main occupation is farming are entitled under Reg. 797/85 to investment benefits; and under Directives 286/72 and 1361/83 to compensatory benefits, subsidies, low interest loans etc. 12 A series of incentive and investment laws (742/77, 849/78, 1116/81 and 1262/82) have been introduced for those rural areas most in need. As were the Integrated Mediterranean Programmes and EC reg. 797/85.

148


I don’t know why women do not want to go out to the farms and do basic tasks such as irrigation. I have not seen one single women getting up in the morning to go to change the water since all of them drive. This could be because their husbands do not want them to go. My husband used to say “my wife does not need to go to the farm. The men like to find everything ready for them when they come back home.” (Maria-widowed farmer, Agia Triada, September, 1996). The farming activities of an elderly work force have been linked with high fertiliser usage (Fanariotu and Skuras, 1991), a reluctance to invest compared with younger age groups (Sakellis, 1985) and less flexibility in farming procedures (Interview with director of Argolid Young Farmers). The progressive ageing of the farming population in Greece has been attributed to the movement out of the industry by younger people. This has occurred partially because of the draw of urban employment and, linked to this, the low status that is often conferred upon farm work. In the Argolid Valley, but not in the periphery, these trends are evident although, because of the location of the valley, they have not coincided with a general depopulation. An additional complication arising from the agricultural age structure concerns the relationship between farm ownership and inter-generational decisions. Policies to reduce the age of the farming population (i.e. EC 797/85) have set limits on the land owned to qualify for subsidy etc. However, the minimum amount of land to qualify will often be passed from father to son with the balance remaining under the control of the father. This balance is often the predominant share of the farm and as such retains control with the older farmer and in so doing contravenes the stated aim of the policy. It may also restrict the options open to agriculture in the region because older farmers tend to be more conservative in their approach. Where work is obtained outside of agriculture then the person will invariably either continue to live in the area or maintain frequent contact from a base in Athens. One model of inter-generational response to farming is outlined by the Proedros of Nea Tiryntha “The percentage of children going into farming decreases continuously. They try to find another job or to study. Suppose that in a family there are three children, the one will stay and the other two will do something else. They are afraid that there will be a crisis in farming. The other thing is that it is a hard job and nobody wants his children to work hard”. This is a very different interpretation from that which is evident in the peripheral villages to the west of the plain (zones one and three). In these areas the majority of children go into farming, rather than into education or sedentary work, and as such they contribute to the maintenance of a more diverse agriculture. For example the following extract from an interview with the village secretary from Statheika varies noticeably from the one above. “All of the young people stay in the village. In these villages the children only go to high school, they believe the less you are educated the better. They think that if the child goes to university he will go away from the village and will abandon the farm”. Those policies that have been introduced to redress the imbalance in the age pyramid have not taken into account the influence that social structures have upon the range of crops that can be supported. 13 Neither have they been directed at those areas that are not suffering depopulation but are


149

experiencing a degrading natural resource that has occurred as the result of farming activity that is heavily water dependent and unable to adapt because of the social, cultural and demographic makeup of the agricultural work force. Central to this, in some areas of the Argolid, is an ageing farming population, but one that has generated sufficient income to enable their offspring to leave the industry. This has restricted their ability to adopt more diverse and labour intensive crops in the future. It is a self reinforcing feature of this scenario that cropping options which are more labour intensive will often be less attractive to the young. PUBLIC ADMINISTRATION, PLANNING AND THE TRANSFER OF INFORMATION The final ‘structural weakness’ referred to in the introduction to this chapter concerns the role of the bureaucracy or public administration in agricultural extension through the monitoring of policy and the transfer of information. It is to this that we will now turn our attention. The Greek farmer is often unsuited to the entrepreneurial requirements that are explicit as well as implicit in a range of the relevant acumen for seizing investment policies.14 This is not just a case of acquiring opportunities, but it is also related to the possession of: “considerable amounts of technical and economic knowledge that would allow them to exploit the opportunities of this new policy”. (Lianos and Parliarou, 1986) While some reservations have been expressed about the adoption of an agricultural model which is based upon concepts of economic efficiency that are culturally inappropriate, the provision of information about land-use options and farming practice is important. It is also perceived, by farmers and state agronomists alike, as central to the role which should be played by public administration along with the collection of data about changes in land-use and their environmental impact. Both aspects of information transfer underpin a need for planning which is seldom perceived to have been realised. This failure to collect and distribute information and to plan on the strength of it is seen to be one which is entrenched in the Greek culture. This is evident both in the academic literature (see Herzfeld, 1992) and amongst the local population. This section will discuss the role played by public administration and the associated failure to plan in agriculture, in the following terms: • Pride in dealing with crises rather than planning to avoid them • The indivisibility of politics from public administration, i.e. through politically motivated staff changes • The population using the constraints imposed by the bureaucracy, real or otherwise, to excuse local inefficiency (Herzfeld, 1992) • Reduction in agricultural extension activities because of understaffing and loss of trained staff • A failure to record accurately and to store information properly

13

E.g. Reg. 797/85, superseded by reg. 2328/91 on Assistance to New Farmers. The requirement to draw up economic investment plans to qualify for support under reg. 797/85 is one example, (see also Tsapatsaris, 1986) 14

150


• Inadequate liaison, and an unwillingness to pass information between agencies. Pride in Dealing with Crises Rather than Planning to Avoid Them The progressive degradation of natural resources in the Argolid, prior to 1995, was generally accepted as being caused by intensive, irrigated agricultural activity. Considerable confidence rested in both the scientific and farming communities about the ability of technology to respond to this situation. This was accompanied by a general acceptance among farmers that the behaviours leading to degradation would continue until the activities involved were no longer possible, or until a technical response was found. Such a short sighted approach is nested in a culture which can attribute the existence of problems to bad luck in oneself but to character flaws in others (Herzfeld, 1992). This is an important synthesis because it introduces the concepts of blame and ‘outsiders’ as integral to a problem and as such creates a role of scapegoat for the public administration. It also raises questions about the cultural bases for planning, some of which are introduced in the following extract from an interview with an agronomist from the Service of Land Improvement (YEB). “It is a question of how a society is educated, and whether they have learned that there are alternative strategies for survival. Japanese society is taught how to respond to earthquakes and they are prepared for them when they occur. I doubt if we have this type of society. In Greece when we face a problem we must leave it until the very end and the situation is hopeless before we look for a solution. We do not know how to plan, our society is such that we cannot anticipate situations and plan for the future. The solution is seen in the crisis itself. The Ancient Greeks said that war is the father of everything”. What appears to be suggested here is a fatalistic approach to managing change, and one that is borne more out of conflict than planning. The public administration can be seen to fit into this model as ‘outsiders’ who can be blamed either for their actions or for their inactivity. This is reinforced by the relationship that has developed between the administrative and the political systems and emphasises the difficulty of establishing a planned framework which can enable the pursuit of desired (sustainable) pathways based upon co-operative effort. The Indivisibility of Politics from Public Administration The link between politics and public administration is a close one in Greece. Tsoucalas, (1991) argues that politics and thereby public administration does not conform to the western liberal model because it has not been borne out of the traditional capitalist economy and the supremacy of the individual. In the Greek context the development of the liberal state apparatus was not built upon the ‘rights’ of the individual so much as within the context of traditional cultural patterns. The state placed Greece in a global context and thereby provided groups and factions with a much wider socio-economic environment. These groups and factions which had previously operated within a restricted local environment were supportive of the state’s position in return for the continued participation in these networks with reciprocal local returns. In short, patronage which was the basis for local socio-economic relations, was extended into the political arena of the state. This provides one explanation for the failure of certain policy mechanisms introduced through the European Union using a western liberal model of public administration. This model assumes a


151

separation between political representation and administration, and for better or worse, such a split is not apparent in Greece. A similar scenario has already been highlighted in terms of the political motivation which has been perceived to underlie decisions relating to water distribution and the allocation of state air mixers in the Argolid. “The Argolid has one big problem, water. It is used politically by PASOK and New Democracy, everybody tells us, I will bring water. People have not revolted about water because it is a political question, since they are for ND and ND is in government they say it’s okay. It was the same before with PASOK, the deputy went to Fychtia and said you will not protest about Anavalos. When ND got into government he went to the village and asked them to protest for it”15 (YEB, Agronomist). By its very nature patronage is a two way process with political contacts and relationships also being developed by local representatives for specific ends. The cultural environment within the public administration therefore allows for direct or indirect patronage in terms of enhanced occupational prospects or the overcoming of administrative inconvenience i.e. obtaining licences and clearing fines. Of course the converse can also be the case with promotion being blocked and licences delayed. This ‘political’ approach to local decision making has two significant effects. Firstly, by reinforcing local differences it prevents any form of integrated planning across the area or through time. Secondly and related to this, is the impact that such a culture has on the public administration. This is manifest in one of two ways, through the active support of a specific political perspective with the accompanying risks should the power base change,16 or through an entrenched ‘bureaucratic’ position based upon ‘non-decision’ making and the avoidance of responsibility. Efthinofovia (fear of responsibility) is perceived to be endemic in Greek public administration. It is also the way that the lower administrative levels in turn justify their inactivity by offsetting blame further up the hierarchy. “While disgruntled clients blame bureaucrats, the latter blame ‘the system’, excessively complicated laws, their immediate or more distant superiors, the government” (Herzfeld, 1992). Herzfeld argues that we must be careful in taking the criticisms of public administration too literally. He cites a tendency to blame outside agencies rather than to accept that internal conflict (village or family) has occurred. Therefore, public administration has been attributed with a dual, and apparently ambiguous role. On the one hand it can be seen as the scapegoat that refuses to accept responsibility, and on the other as the basis for personal patronage and the removal of bureaucratic obstacles. Indeed the latter role was at the heart of the high status accredited to employment within the administration and was one reason for family/community support for such employment albeit at low rates of pay.

15

Communities tend to have distinct political allegiances, in some of the larger villages these allegiances may differ with the groups congregating together, usually in the cafes.

152


Reduction in Agricultural Extension Activities The recent increase in agricultural production has become subject to an increasing amount of monitoring and control i.e. production monitoring for price support and quality control. This has coincided with a loss of public confidence in the public administration which is indicated in general terms by the reduction in the number of people wishing to enter the administration, and in the case of agriculture through a decline in the use of the administration for advice. This loss of confidence has partially occurred in response to public scandals,17 but more particularly it has been the result of staffing reductions and the subsequent difficulties in undertaking wider responsibilities. As a result the role relating to the provision of information and agricultural extension is perceived by the farming community to have been neglected. This in turn has led to a demoralised work force that is perceived to be more concerned with self preservation than with service provision and as such does not have the trust of the farming community.18 A more charitable view also exists which is that the decline in agricultural extension is due to understaffing and to the increased work load of those agronomists who are in place.19 “Between 1975 and 1981 we had an agronomist twice a week. He came to the coffee shop and anybody who needed him could take him to their farm. The situation has changed because the service has a lot of work and limited staff”. (Farmer, Elliniko) “I cannot find the local agronomist. They are not to be blamed because they are responsible for many things; sheep, olive oil, exportation. We have always requested an increase in the number of agronomists. I will give one example of what this can mean. We declare our sheep and goats to take the EC subsidy and have not yet taken last years money because there has been nobody in the service to work on it.20 I will tell you something else, each party which comes into power dismisses personnel and after a certain period replaces them with ‘their own’ agronomists who do not have local knowledge”. (Farmer, Karya) It would appear that the presence of agronomists in the community and on the farm was one vehicle for overcoming the scepticism which is attached to the public administration as an entity. While this is reflected in some of the legislation intended to support extension services,21 in reality the contrary is perceived to have occurred over the past ten years. A process that is compounded by the additional responsibilities incurred in order to comply with EC requirements. While the monitoring role has increased there is concern that the information transfer and advisory roles have declined at an even more rapid rate. This has resulted in farmers having to actively search for information about new policies and legislation. For example, in response to the fruit restructuring programmes in the late 1980’s

16

There are frequent stories about agronomists with a known political view being moved to isolated locations or to lower status jobs after the election of an opposing party. 17 For example over olive oil production. 18 E.g. there is often a requirement for agronomists to ‘clock on’ at the required time even though this inevitably fails to coincide with the working timetable of the farming community.


153

“Nobody came from the Service of Agriculture to tell us these are the programmes and this is what you do. At the time subsidies existed but the agronomists were employed ‘policing’ the dumping of oranges” (Secretary of Perseas co-operative in Argos). This situation was perceived to reinforce an inequity within the farming community by which those farmers who lived in villages away from the urban centres, and thereby the administrative offices, were not informed about policy changes and the introduction of new programmes. This meant that many benefits were accrued by those who were not seen to be ‘real farmers’ but who were centrally located and may have “personal contact with an agronomist or somebody in the Service of Agriculture who was responsible for the programme” (agronomist). This problem is accentuated by the fact that those who farm away from the central plain are often less educated, working full time on their farms and as such less able to negotiate their way through the administrative procedures. This in turn is exaggerated by the common perception among farmers that the Service of Agriculture agronomists are not qualified and act purely as clerks. “They offer no advice about the cultivations and we only use them to sign certificates. The dealers of pesticide do show an interest because they want to sell” (Farmer—Argolid Plain).

This latter point represents an important transformation in agricultural activity and the qualitative condition of natural resources. Pesticides were previously distributed through the Agricultural Bank, they are now sold under licence, by qualified agronomists who also act in an advisory capacity to the farmers. This can be seen as a conflict of interests with results that are inevitably detrimental to the environment. A reservation that is expressed by a farmer in Pyrgella who is actively attempting to introduce biological farming methods. “The agronomists are dealers, they want to sell. When a new pesticides appears they know that if they sell it their percentage profit will increase”. This farmer continues by explaining that the depressed state of agriculture in the area at the moment means that farmers will panic at the onset of any disease and will rely upon the agronomists to provide information and products to offset the perceived risk. Indeed, the agronomists themselves may be poorly informed about ‘ecological’ farming. “They give chemicals to kill the weeds, but they don’t think that maybe these weeds have a useful role to play in the ecology of the farm”.

19 In 1992 there were three agronomists instead of the allocation of seven responsible for the whole of the Argos district. 20 Agronomists from the Service of Agriculture have to verify claims and process the relevant forms for the EU. 21 Reg. 797/85.

154


Therefore, the introduction of registered agronomists in a self employed capacity has placed a resource closer to the farming community, but equally one that has a vested interest in the sale of certain products, such as fertilisers and pesticides. This raises the question about who the farmers would approach, and listen to, for information about farming options. The Ministry of Agriculture have no direct contact with the farmers, whereas the local agricultural service, sometimes through understaffing, has been seen to assume a bureaucratic distance and a reluctance to assume any form of responsibility. This has generated a disillusionment that affects the information, knowledge and thereby options available to the farmer. It also represents a reluctance on the part of many farmers to trust advisors from the public administration, particularly when they also operate in a monitoring or policing role. As a result, the private agronomists and possibly more importantly, other farmers are the most frequent source of information. Failure to Record Accurately and to Store Information Properly The problem of understaffing within the public administration provides one reason for the lack of extension services that are provided. It is also a contributory factor to the inadequate information about agricultural production systems that is transferred from the community to the political decision makers. This deficiency is evident in two ways, firstly in the quality of the data that is actually collected. One agronomist recounted that a farmer had entered two elephants on his livestock returns for over a decade before it was noticed. Secondly, the process by which data is correlated, stored and transferred is often inadequate. This follows on from the perception among the farming community of the agronomists operating as clerical staff. In the Nafplio Service of Agriculture records are manually transcribed by the agronomists, a procedure that is time consuming and inevitably constrains their ability to collect data and advise on site.22 This can also be used as a way of restructuring a work force, either by moving ‘dissidents’ into desk based work or by frustrating them to the extent that they leave. In the same office limited information technology was available, but the only trained operative had been relocated. Such organisational explanations of poor data collection cannot be separated from what has already been discussed as an unwillingness to plan, political involvement in public administration and a reluctance to share information. The agronomist working for the Land Improvement Service makes the point that “nobody knows the exact information about anything here in Greece because we do not know how to count, to measure and when we do measure we discard it.” He continues that even when information has been properly collected it is not catalogued or stored properly. More commonly, data is thrown away and collected again when necessary. As a result there is nothing to build upon. These constraints also affect the capacity to undertake research about the relationship between the agricultural system and the natural environment. The organisational parameters of the administration were defined by stated objectives and the status quo and this meant that “YEB detested every form of research” because it could question that status quo. The agronomist continues that most research that was useful and innovative was undertaken 22 Information about production is collected by village secretaries (employed by the Ministry of Internal Affairs and the OGA) in liaison with the farm policeman and a village committee. This information is provided annually with a six month update to the Service of Agriculture.


155

through the personal interest of the agronomist concerned “with no help from anybody, no instruments, no driver etc.”. When research was undertaken by agencies external to the public administration, even if this was organised by local farming groups,23 there was some concern about the accessibility of the information provided. This was seen to stem from two factors, firstly the language used was that of the ‘expert’ and as such perceived to be elitist whereby farmers were not informed because the Professors who spoke did not want to inform them. As a result even the limited scientific information that is obtained does not necessarily reach the public domain. “It is not expanded and is considered the property of the scientist” (Interview with representative of Argolid Young Farmers). Secondly, there was felt to be considerable rivalry between experts and agencies and this greatly hindered the free flow of information. One final point which is related to the poor communication between agencies, and between agencies and the farming community, is the dispersed responsibility for policy implementation. For example the promotion of agro-tourism24 was not the responsibility of any individual agency and as a result was perceived to be poorly co-ordinated because each agency was struggling to gain authority over the others rather than meeting the objectives of the policy. Consequently farmers who may have been interested in the programme did not receive the necessary information, lost heart and on occasions returned to traditional and established but vulnerable crops. STRUCTURE AND CULTURE: CONCLUSIONS This chapter has looked at the structural characteristics of Greek agriculture alongside the situation encountered in the Argolid Valley. It is apparent that what are seen as structural deficiencies in one part of Greece are not necessarily so elsewhere in the country. Similarly, what is considered to be economic efficiency in Northern Europe may not be applicable to the cultures of the Mediterranean. The history of land ownership, multiple job holding and the relationships between the community, political representation and public administration all contribute to an agricultural structure that is not easily assimilated into such a system, or indeed would benefit from being so. A number of points were also raised that were specific to the role played by the public administration and its relationship with the farming community. • There was perceived to be a pride in dealing with crises rather than in planning to avoid them • Politics was seen as indivisible from public administration • There has been a reduction in the impact of agricultural extension activities because of understaffing and loss of trained staff • There was a perceived failure to record and monitor accurately and to store information properly

23 I.e. The young farmers of the area organised an Hemerida (seminar) on the subject of declining water quality and stocks. 24 Under regs. 797/85 and 2328/91.

156


• Inadequate liaison between agencies, and an unwillingness to pass information between them, has hindered the successful implementation of policies. These issues will now be considered alongside the perceived uncertainty attached to farming in the area and will subsequently be linked to the bio-physical and agricultural economic factors discussed in earlier chapters to move towards a crop choice framework. This in turn will be used to inform, and interpret the output of, the modelling activity undertaken for the study.

9. PERCEIVED UNCERTAINTY AND FARMING: ESTABLISHING A FRAMEWORK FOR CROP CHOICE Mark Lemon and Nenia Blatsou

The previous three chapters have explored the factors that influence the type of farming carried out in the Argolid Valley. These have been classified under the headings of the bio-physical characteristics of the land that is farmed and the condition of natural resources on that land, the technology that has been employed to support agricultural activity and the socio-economic characteristics of agriculture in the area. Attention has also been given to the institutional arrangements that are in place to manage different aspects of agricultural activity (i.e. co-operatives, local and central government). What has not been fully considered is the uncertainty that is perceived by farmers and the behaviours that they anticipate in response to that uncertainty. This chapter will draw upon data obtained from the field work interviews with farmers about the conditions of risk and uncertainty that pertained in the early part of this decade (i.e. low rainfall, degraded resource base, levels of price support) and will update these in the context of current concerns (i.e. future of price support and market competition).1 This analysis will then be combined with that of the previous chapters to establish a framework within which cropping decisions are made and to put forward a provisional taxonomy of farmers and their decision space. UNCERTAINTY ABOUT NATURAL PHENOMENA The initial exploratory interviews with farmers established rainfall levels, frost, and pests and disease as the most important sources of concern about natural phenomena as they affect farming behaviour. It was intended that a spatial distribution of perceived threat be established throughout the valley and that some insights formed into whether or not farmers saw these threats increasing or diminishing in the near future and their likely responses to these scenarios. If we look at perceived natural threats across the study area it is apparent that frost and decreasing rainfall were predominant in 1993 when the original research activities were undertaken (Table 9–1). When this distribution is seen in terms of the zones it is clear that farmers in the coastal area of zone six were preoccupied with the problem of frost whereas those in the peripheral zones three, four and seven, and in zone two, were more concerned about any reduction in rainfall. With the exception of zone five this coincided with the qualitative and quantitative degradation that was apparent in each of the zones. One explanation for the distribution in zone five could be that the northern and eastern parts are more prone to water shortage than the south which suffers primarily 1 Current concerns were elicited from semi-structured interviews carried out with farmers in 1997. These were undertaken to investigate how farming conditions had changed in the area since the initial research in 1993 and to provide an indication about the relative importance attached to different phenomena.

158


Table 9–1 Concerns about natural phenomena distributed by zone.

from frost. This was exacerbated by a greater use of bore-holes than in zone six and as a result the need to leach out additional salts from the soil. Zone one provides an interesting spread of perceived natural threats that is indicative of its topography with some frost on the coast and the threat of water loss in the hills. The perceived threat of crop virus and disease was generally not considered to be as significant as frost and rainfall although in zone one it was cited four times. The increased variation in crops in this zone means that it was more likely to experience some form of crop disease. Indeed the area was particularly badly affected by the Sharka virus upon the apricot crop. Paradoxically, because of this variation in crop the area is less vulnerable to any single virus or pest than the areas that are predominantly single crop, zones five and six. Two hypotheses can be drawn from this observation. Firstly that there was less concern about the dangers of pests and virus on citrus crops2 because they had not been recently experienced in these areas and secondly, these threats were not comparable to that which was perceived to exist from frost. The late frosts of the last three years have led to the loss of apricot and orange crops in the peripheral zones. This has dissuaded many farmers from cultivating Clementine mandarins which are perceived to be economically beneficial but vulnerable to frost and therefore requiring a high capital investment in frost protection. With the exceptional levels of rain between 1995–97 the issue of low rainfall has become secondary to that of frost, particularly in the coastal areas. As was seen in chapter seven considerable attention is now being paid to flood management, although this is not restricted to the agricultural community and is particularly relevant in the urban areas adjacent to the coast. The perception of disease and viruses as a threat to agriculture remains low with initial concern about the arrival of Aleurotrixus floccusus and Phyllocnistis citrella being countered by confidence in the capability of scientific solutions in the form of biological and chemical treatments. PRICE, UNCERTAINTY AND RESPONSE The reduction, leveling off and removal in price support from some crops (i.e. olives, oranges and mandarins respectively) combined with frequent delays in payment have created considerable economic uncertainty in agricultural production of the Argolid. This is particularly important in the case of the Merlin orange which has dominated agriculture in the main valley, been considered resistant to pests and assured of a market. The situation has become markedly more serious over the

2

E.g. Coryphoxera on oranges, although this had previously decimated the local lemon crops.

PERCEIVED UNCERTAINTY AND FARMING

159

last two to three years as price support has dropped dramatically (see chapter six) with a corresponding decline in confidence in the main valley which previously has enjoyed considerable external financial support. This is less apparent in the peripheral zones where a more varied agriculture is practiced (i.e. citrus varieties and access to different markets). Concern is therefore less apparent in these areas about the reduction in price support and indeed it is often seen to be a positive trend encouraging all of the farming community to compete in a free market arena. Therefore the inequity previously perceived by the peripheral farmers under subsidised conditions has been reversed with those from the central plain appearing less able to adapt to the changing situation (i.e. unable to commit the necessary time). When asked whether they felt that the price of their main crop would increase or decrease in the forthcoming year (1994) half of the respondents (44%) felt that there would be a small increase and a similar number felt that the price would stay the same or decline slightly (24% each). The remaining 5% felt that prices would decline significantly. This data fails to differentiate between the type of market that is targeted by each farmer. An analysis of the spatial distribution of these responses did not indicate any marked variation between zones and as such would suggest that this distribution is applicable to both the local and supported markets. The situation in 1997 has changed considerably with no farmers in the main valley expecting the price for their main crop (Merlin oranges) to increase. This is because they are tied into a price support framework which is being progressively lowered by the European Union. Alternatively farmers from the periphery who sell to the local markets are more likely to anticipate increased prices even if this is only to keep pace with the inflationary trend of the drachma. The state of the local economy has also benefited the farmers who sell to the open market because they receive payment immediately and therefore have capital available to be spent locally. Alternatively those who sell through the co-operatives and dealers under price support have to wait for payment which means the value of that payment has reduced accordingly. Whereas in 1993 there was a mixed anticipated response to concerns about price, the current situation has resulted in many farmers from the central valley reducing the level of inputs (fertilizer, pesticides). This reduction has been accompanied by a growing level of concern about the health issues surrounding the excessive use of pesticides and fertilizers. It appears, however, that the driving force behind this reduction of inputs has not been the health implications but the reduced income from agricultural produce. In the peripheral villages where this drop in income has not been evident there also appears to be less concern about the health issues attributed to farming practice and input levels have not changed noticeably. ‘The restructuring was not accepted by the farmers. They thought, ‘why should I cut down a tree that is thirty years old?’ When I uprooted my apricot trees because ofSharka I felt so sad that I left for one week. I had seen them grow like my own kids’. ‘Generally I don’t listen to what somebody else tells me to plant on my farm. I plant whatever I like’. (Interview with farmer from Hera). FIGURE 9–1 ATTITUDES TO CROP CHANGE IN 1993.

Farmers were also asked whether they would contemplate a change in crops in response to the anticipated price change. In 1993 very few respondents contemplated changing crops in response to

160


dropping prices, this is not the case in 1997 with many farmers in the central area being prepared not only to change the varieties of citrus that had been previously rejected (Roco, Salustiana, Newhall) but even to contemplate selling their land. This raises an important methodological point which is that the presentation of scenarios can elicit very different anticipated behaviours from those which may be considered in response to changes in actual conditions. The following statements, and variations upon them, were common in the interviews undertaken in 1993. Indeed some of the implications arising out of them have already been considered in terms of access to, and use of, sources of information and the related problems of planning for the area. The second statement conveys the individuality of the farmers rather than being a representation of arbitrary decision making. The issue surrounding the provision of information has, however, become far more acute for farmers in the central areas with the decline in income and an increased willingness to change their farming activities. (See Figure 9–1) Complaints about the poor communication of policy changes from the Service of Agriculture are even more prevalent and are often used to explain the reluctance to change crops. Previously this reluctance was more related to a stated affinity with the crops and perhaps more importantly a reluctance to ignore the considerable investment required to establish perennials. One aspect of that investment is the time required for perennials to mature and the other refers to the expenditure to install the necessary infrastructure (irrigation systems, air mixers etc.). Obviously certain annuals also require considerable investment (greenhouse crops), however the take up of this type of farming is limited in the Argolid. More prevalent is the open cultivation of annuals in the peripheral areas. These are relatively easy to move between and do not require such heavy capital investment. They do however, have high labour requirements and as such are unsuited to the central valley where farmers invariably have other occupations, limited time and often a reluctance to undertake more intense physical farm work. Certain crop changes have been forced upon farmers as a result of plant diseases and viruses i.e. Sharka on apricots and Coryphoxera on lemons. In the case of the Sharka virus separate support was available for the uprooting of the apricot crop and its replacement by suggested crops. The latter has already been seen to fail because of an inability to relate the suggested options to local structural conditions and the cycles of agricultural production (e.g. the delay between the end of the subsidy and the maturation of the suggested crop).3 This resulted, in the short term, in increased tobacco production, and in the longer term, in the planting of citrus trees or even the replanting of apricots. TOWARDS A FRAMEWORK FOR CROP CHOICE The preceding chapters and the earlier part of this chapter have examined the bio-physical and socioeconomic characteristics of agriculture in the Argolid Valley. It has been argued that farmers make their decisions upon the world as they perceive it however there are certain features which constrain their ability to meet the objectives associated with those decisions (e.g. extreme climatic conditions which jeopardise their crop). Alternatively there may be potential facilitating factors that may not be known or appreciated (e.g. inaccessible information about improved farm practice or potential financial support). Similarly the example of technology could be used to differentiate between opportunity and decision space whereby the existence of a technology could fulfill a local need, for better or worse, but a lack of information about it would mean that it lies outside of the decision space of the farmer. There is therefore an opportunity space within which farmers could potentially


161

Table 9–2 Dimensions of farming opportunity space.

make decisions and a decision space within which those decisions are actually made. Table 9–2 presents the basis for a crop choice framework which pulls together those dimensions that have been seen to influence the (in)ability to farm in a particular way in different locations. This will then be used to form the basis of a provisional typology of farmers which can inform about potential decisions and responses across the valley. The term ‘dimensions’ has been chosen specifically to indicate a range within each attribute as well as the possible combinations between them. However it is not suggested that they have equal weighting or the same characteristics in different locations. For example there are a range of soil type and levels of water quality and the social acceptability of farm work will vary across communities. Figure 9–2 provides a schematic representation of how individual and collective interpretations (actors A,B,C) of a situation do not cover the total possibilities that are available. Obviously questions relating to information flows and knowledge creation are important here. For current purposes however it is sufficient to suggest that decisions have multiple influences and are not constrained to bio-physical or economic parameters. Indeed it is possible for individual decision makers to perceive options that do not exist. For example, this may occur as the result of mis-

162


Figure 9–2 A schematic representation of opprotunity and decision space.

information about policy innovations such as new subsidies. In such circumstances it is conceivable that part of the decision space within which an individual or group are operating is actually outside of what is possible. The multi-dimensional nature of opportunity space presented in Table 9–2 means that different decisions are likely to be made, both in different locations, and among the same population at different times. At the same time it would not be useful to present a framework that treats each farmer as a unique entity operating from an idiosyncratic position. Rather some commonality needs to be identified in order to support a policy relevant analysis. This is of particular importance with the more abstract dimensions of the framework referring to socio-cultural factors (i.e. status attached to agriculture) that cannot be readily entered into a computer based model. Indeed it is only through the contribution of social science and local actors that the cultural environment within which decisions are made can be incorporated into the modelling activity. This is an important point that needs to be borne in mind when considering the final section of the book on computer modelling. Before embarking upon an introduction to the modelling activity it will be useful to draw together some of these dimensions and suggest a provisional typology of farmers operating in the Argolid Valley as a template through which the modelled futures can be interpreted and modified. It is also important to note that such a template is not static and to this end the typology developed for the situation that was observed in 1993 has been modified to accommodate the changes that had occurred between that time and the 1997 study.

3 Nobody in N. Tyrintha planted the suggested crops of pistachio or trifolium. This was largely because no information was provided to the farmers about production and income levels. Elsewhere the lack of peer experience and the temporalities of suggested crops were cited as reasons.


163

TYPOLOGY OF FARMERS Group One—1993 The first group are predominantly mono-croppers of irrigated citrus trees in the central zones five and six with high productivity per unit and the adoption of the European Union price support system rather than local markets. A high level of technology is employed to offset the effects of degradation and frost. Primary employment and investment are outside of agriculture and the ability to farm in a more diverse way is constrained by limited family and personal commitment to working the land. The low status accredited to farming within this group is further exemplified by the tendency for farm children to enter higher education and move into occupations other than farming. Due to the existence of second incomes this group had a high price flexibility which meant they were able to withstand price reductions and as such were less vulnerable, in economic terms, to degrading natural resources. They were however less flexible with regard to labour intensive cropping and marketing options. Group One—1997 The dependence of this group upon the Merlin variety of orange has meant that with the continued reduction in price support for the crop and the uncertainty surrounding payment they are now considering the sale of land and change in the variety of citrus they grow. They remain constrained from more labour intensive farming by their commitment to other occupations and an unwillingness to personally engage in manual farm work. In addition there is a high technological component to farming among this group and as prices have dropped the high costs attached to that technology have become out of proportion to the income generated. Therefore, where this group was relatively stable and self-confident in 1993, with little perceived uncertainty, they are now considered to be in an extremely vulnerable position. Group Two—1993/1997 The second group of farmers within the typology are situated in zones four and seven on the northeast periphery. These generally have larger farms, due to the landscape, with limited crop variety and low levels of technological investment to combat the effects of degradation. There is low productivity due to the hilly terrain, water degradation and poor soil in the areas that they farm. These farmers are therefore constrained by the need to generate additional income outside of agriculture, a factor which also restricts the potential for children to enter farming. The external income (i.e. through tourism in Mykenes) can become essential to the household budget and thereby lead to a further decrease in the time allocated to farming. These factors have restricted the ability of this group to adopt cropping options which require high levels of labour or reliable, good quality natural resources. Because of the bio-physical conditions under which this group farm, and the fact that they had significant olive cultivations and were not totally entrenched into the artificial market, they have avoided much of the significant change in the area. The farmers in this area remain dependent upon other sources of income even though many of them would prefer to farm full time.

164


Group Three—1993 The third group are based primarily in the west of the study area (zones one and three and the western part of zone two). They have a greater diversity in the type and variety of crops grown and the markets to which they sell. A high level of agricultural investment is made both through personal (family) commitment and expenditure on technology and farm based investment. This commitment is manifest through the supportive role played by the extended family and the tendency for children to enter farming on leaving high school. There is considerable flexibility in the type of farming undertaken by this group due to the level of commitment to farming, the use of ‘hidden labour’ for marketing and work in the field and the domestic consumption of some crops (vines, cereals etc.). There was some concern in the early 1990’s about the equity of local agricultural structures which benefited part-time farmers in the central area who were not perceived to be authentic or real farmers, but who received the bulk of agricultural price support and subsidy, had access to the Anavalos infrastructure and yet were seen as the cause of much degradation by over exploiting the aquifers. Group Three—1997 This group have become more confident with the recent high levels of rain removing much of the concern about water depletion and restricted access to Anavalos. They continue to adopt a polyculture system of farming and to sell to the local markets with the cash flow and capital benefits attached to that choice (see above). Group Four—Young Farmers A final group of farmers are clearly identifiable in the Argolid and perform a role over and above the working of their own land. The Association of Young Farmers has developed over the past decade, primarily through the children of farmers in the central area who entered tertiary education and decided to return to full time farming. In some cases this education has been in economic rather than agronomic related disciplines. This has enabled the farmers concerned to consider other options, such as greenhouse flowers, where additional agronomic knowledge is considered to be of less importance than economic management (Blatsou, 1996). This group question existing practice, particularly that which is based upon high input monocropping, and have advocated a more sustainable approach which adopts a longer term perspective. Consequently they have very different attitudes towards farming to those held by many of their parents who would be classified under the first group of this typology. The Young Farmers tend to combine this reduced input approach with a sophisticated use of information technology. This enables them to keep in touch with developments in agronomic research, to establish and maintain contact with other farming groups across Europe and to update information about agricultural policy (i.e. within the European Union). As a consequence this group are a source of information to other farmers in the region, both on a personal basis and through the organisation of seminars on a range of issues salient to the farming community (i.e. water quality, policy changes and implications). The discussion in this chapter has highlighted two factors that are relevant to land-use policy. Firstly there is a requirement to take account of the system as it is perceived by the local actors and secondly a need to consider configurations of attributes that are not defined by any particular system


165

i.e. natural or social, environmental or agricultural. It is the ability to interpret local change as an evolving phenomenon within and between sub-systems which provides us with a starting point for defining desired states that are emergent trajectories or paths, rather than specified end states (Park, 1993). This is fundamental to sustainable land-use because it recognises that what is sustainable continually changes according to the social and physical circumstances in a locality, as well as the knowledge base of the local actors. These variations are affected in turn by external changes such as the level of crop support prices or the emergence and adoption of technological innovation. The combination of these two factors—the world as it is perceived by decision takers and the complicated configurations of factors that determine that perception—cannot easily be represented through a computer model. Decisions can be ‘switched’ on and off according to price and water availability or constrained by poor soil or absent markets, and in more sophisticated models these factors may be taken linked together. It is however the concurrent and changeable configurations of factors (dimensions) that provide the ‘rich’ and emergent picture through the eyes of decision makers. While claims to represent this must be treated with caution, the selection and weighting of attributes to be included in the model, and the interpretation of output, by decision makers or agents capable of representing them supports an iterative process that is beneficial to the authenticity of the model. It is with this interpretive framework in mind that the final section of the book, concerning the modelling activity, should be read.

166

10. POLICY RELEVANT MODELLING IN THE ARGOLID: FROM SOCIOLOGICAL INVESTIGATION TO CROP CHOICE MODEL Roger Seaton and Ian Black

The final section of this text will pull together the story line that has been introduced in the previous four chapters and will consider how computer based models can be used to represent that story in a way that is policy relevant. Methodologically this section will also highlight the need to integrate different types of model which represent data at varying levels of resolution. The data used for the models reported in the following chapters refers to the research phase ending in 1994. Whilst this means that more recent events such as the increase in rainfall and aquifer replenishment will not be represented, the techniques employed remain relevant This chapter will present a model of crop choice and chapter eleven will describe the background to modelling the water/salt system. Ongoing attempts to develop an integrated strategic representation of the system will then be presented and a concluding discussion will consider the policy and methodological implications of the work. More specifically the tasks in this chapter are: I. to contribute to an improved understanding of the linkages between policy instruments and the receptivity of farmers to these. II. to link the sociological data presented in the previous section to the model of farmers’ crop decisions. III. to outline the crop decision model. As was described in earlier chapters a number of significant changes have occured in farming in the Argolid over the past fifty years. These relate primarily to the agricultural water system and the economic dependency of the area on farming, both of which have caused concern to the farming community, scientists and agronomists, and in policy circles within Greece and the EU. The simple extrapolation of these changes into the future has resulted in a number of “issues” being articulated as “problems”. The starting point of the approach adopted for this study is that change is a normal condition of any complex situation in the real world. From an anthropocentric position a “problem” always depends on the effect it has on human welfare. Thus “problems” are often articulated differently by individuals and groups according to their perception of existing or future conditions. For example, the farmers refer to the “problem” of water when, in quality and quantity, it becomes more difficult to obtain, perhaps because of price or rationing. On the other hand hydrologists may refer to a “problem” when they perceive the effects of demand for water on the depth and salt content of aquifers. A soil specialist, however, may perceive a “problem” when the soil deteriorates in its ability to support vegetation whereas an agronomist may be primarily concerned with the reduced potential to grow a particular crop. Part of the Argolid research project has explored these different perspectives about what constitutes a “problem” and how these link to

168


Figure 10–1 The physical environment, production and farmers decision making.

current and future human welfare and the decision space within which farmers are operating. It is generally accepted that the current economic welfare of the region is directly linked to agricultural production, and through that production, to the quality of soil and water. One objective of the work has therefore been to model the decisions by farmers about what crops to grow and how to grow them. Therefore the Argolid research has been concerned with the variety of influences that affect the ability of the physical systems i.e. soil and water, to support food production. In particular, the work has aimed to identify the conditions under which there is permanent or very long term loss of options about what can(not) be grown as a result of degraded natural resources with a related reduction in welfare. This highlights the interaction between land and water quality, other aspects of the physical environment and the farmer community (Figure 10–1). It has been seen that farmers in the Argolid have questioned their choice of crops for a number of reasons over the past decade. In simple terms this was caused in the period prior to 1994 by the increased cost of clean water for irrigation. Since that time concern has focused more on the growing uncertainty over levels of price support and the mechanisms of payment. While these concerns have led only to a limited change in crops the implications and costs of transition from one configuration to another are important to understand. The crop choice framework outlined in chapter nine presented a complicated configuration of social, economic, political and bio-physical attributes some of which lend themselves to inclusion in a computer based model of crop choice. Others however are clearly more difficult to represent and may be better employed as mechanisms by which the model can be defined and interpreted. For example, a reluctance to work the land limits the potential for labour intensive crops. This could be incorporated into the model by weighting against certain crops in areas where this attitude prevails. Alternatively the output of the model could be examined for scenarios which appear to be culturally inappropriate. This is an important point not only in terms of how models are formulated but because it emphasises their role as tools for exploration rather than prediction. The purpose of the model(s) was to support this exploration not by incorporating all of the determinants of decision making but by selecting those that could be realistically represented. It is this balance between identifying the salient attributes, representing those that are appropriate within a model and exploring the results in the context of the ‘big picture’ which is central to interdisciplinary policy relevant work. It also highlights the need to incorporate the stakeholders in the identification of attributes and ultimately in the interpretation of outputs. The overall objective of the modelling activity was therefore to select those attributes that could appropriately represent the relationships between economic welfare, food production and soil and

POLICY RELEVANT MODELLING IN THE ARGOLID

169

water quality. Economic welfare was associated with changes in the profit a farmer makes by growing crops whereas the food production process required information about the selection of crops and input use, particularly water. TOWARDS A CROP CHOICE MODEL: THE POLICY DIMENSION The intention of the modelling activity around crop choice was to explore the variables and relationships in this food production setting i.e. what influences and what is influenced by the use of land for the production of food. This part of the project, which addresses economic and agricultural issues as a contributory factor in soil and water degradation, is located with respect to other relevant policy issues in Figure 10–2. The shaded boxes indicate the focus for this chapter. The aggregate effects of agricultural activity in the Argolid prior to 1994 (reduction in water quantity and degradation of both soil and water quality) are of fundamental concern at the policy formulating levels where they could be see to conflict with the objectives of the European Union about welfare and eco-system degradation. The fact that various policy instruments were introduced at different stages to intervene in the situation implies some form of “failure” in the existing state of affairs arising partly from farmers being uninformed or misinformed about, or simply ignoring, the likely effects of their decisions. More importantly the objectives and decision issues of individual farmers was different from those of farmers as a whole. As will be seen there are no mechanisms by which the individual farmer is required to take account of the effect of his decisions on the future of others. The sum of the behaviour of individual farmers results in a situation which is not congruent with their collective long term objectives nor with Greek Government and EU objectives. It is from this perspective that the modelling work has been undertaken and is therefore concerned in turn with identifiable changes in economic “welfare”.

Figure 10–2 The focus of policy relevant modelling in the Argolid.

Policy Instruments Policy instruments are the mechanisms for intervention which influence the relevant variables in this system (i.e. welfare, crop production and land/water quality). The previous analysis identified a

170


range of broad policy instruments which can influence the key agricultural processes in the area. These will be briefly considered in terms of their influence over the attributes incorporated into the crop decision model. Figure 10–3 provides a schematic representation of this model with the range of policy instruments considered identifiable by the numbers referred to below (I n.m). Instruments focused on the farmer • Management support (I 1.1): By management support is meant the provision of technical and managerial information, education and training, the support and advice of professional agronomists and the opportunities for self-organisation and self-help provided by the cooperatives etc. • Technological support (I 1.2): Technological support is interpreted as enabling improved access to production technologies i.e. bore-holes, pumps, air mixers, sprays, filters. Instruments focused on the crop • Price support (I 2.1): Or price guarantee whereby an income is ensured for produce. This is generally organised through the co-operatives although considerable uncertainty currently exists about the levels of guarantee and the mechanisms for payment. • Subsidy (I 2.2): Subsidies do not currently exist as a direct payment to the farmer for citrus crops, although there are limited subsidies for olive oil and cereals. Where crop subsidy has been involved it normally concerns the abandonment of crops, such as apricots, so that farmers are paid to remove infected trees or trees likely to be infected. Although this is seen by agronomists, the Department of Agriculture and farmers as part of a disease control system, it also functions to ease the cost of transition from one crop (in decline or failure) to another. Instruments focused on water • Water pricing (I 3.1): The cost of water is influenced by the technology employed to access it (pumps, drilling etc.), and the electricity or fuel required. Alternatively it may be purchased from neighbours or more likely from the local authority responsible for managing the Anavalos canal infrastructure. • Water Control and Rationing (I 3.2): The Anavalos system also employs a form of rationing whereby water can only be used for certain crops and at certain times of the day. Similarly in some peripheral villages suffering water shortage, but without access to Anavalos, a communal source of water may be purchased from the community but only according to certain restrictions as to the timing or amount of abstraction. Model Framework A framework for a policy relevant techno-economic model can now be suggested. This places the decision making of the farmer about crops at the centre. Thus and the model needs to include


171

and the model needs to include • the objective variables; welfare, food production, soil/water quality • the factors that sociological research has shown to be the most relevant to farmers for their economic activity on the land • the way in which various policy instruments impact directly or indirectly on decision making. The decisions which result in particular configurations of crops and land-use can be seen as a central focus around which to develop a policy relevant model. The schematic of this model is shown in Figure 10–3. By focusing on the choice of crops it is possible to identify the ways in which the relationships between policy, farmer and natural resources can be linked together. Thus it is possible to formally link the effects of policy instruments on farmers decision-making about crop choice to changes in the soil and water systems. We will now move on to consider the crop choice model more formally. MODELLING CROP CHOICE IN THE ARGOLID The second part of this book looked at the technological, socio-economic and biophysical systems which combine to influence the decision space, relating to crop choice, within which farmers operate. This chapter will return to crop choice in the context of how it can be modelled, firstly through a brief discussion of relevant literature and secondly through a description of the approach that was adopted for the Argolid study.

Figure 10–3 Schematic representation of crop decision model.

172


It is clear that an understanding of farmers’ choice of crop is critical to any policy relevant modelling in the Argolid. Crops and their consequent demand for water are central to the policy concerns in the region and any policy initiatives must inevitably be evaluated in the light of their impact on farmers’ choice of crop. A farmer’s choice of crop or mix of crops is clearly based on a complex set of interacting factors. Closely following Perry (1986) it is possible to identify seven factors that need to be considered: 1. The personal characteristics of the decision maker. These include his (or occasionally her) short and long range goals. These goals will almost certainly include a financial dimension but may also incorporate personal preferences towards risk, land preservation, religious considerations and family preferences. 2. Climatological and Topographical Conditions. These refer to the soil types and relevant aspects of the climate (rainfall, temperature, frost days for example) in the area of interest, all of which will influence crop production possibilities. 3. Resource Constraints. These include natural resources (land area, water availability and price), human capital (labour availability, skills and price), physical capital (machinery) and financial capital (availability and cost). 4. Government Influences. These include restrictions on production or area of production, subsidies and financial assistance schemes. 5. Market Conditions. These include the expected emerging profile of prices, their uncertainty and the location of suppliers. 6. Production Possibilities. These include expected yields as a function of soil, climate, resource input and available technology. The interaction between different crops, animals, pests and crop rotation also need to be included under this heading. 7. Other Factors. These include landowner’s preferences (as opposed to the tenant) and creditor’s preferences. All of these last six factors contain two elements—the mean expectation and the uncertainty surrounding this mean expectation. When referring to these factors it should also be recognised that it is the decision maker’s perception of these factors which influence any decisions and these may not correspond to objective measures of these factors over a sample of years (climate) or farms (production possibilities) for instance. The modelling challenge is to incorporate a set of these factors into a formal decision making process that usefully describes the crop mix by individual farmers or, more usually, a group of farmers in a region and the responsiveness of that mix to different factors over time. Usefulness in this context can be defined in terms of its ability to represent actual crop mix in the region and track its response to variables which are of interest to policy. In terms of decision makers’ goals the simple objective of profit maximisation has received most attention. Bellman (1985) describes a model which assumes that farmers maximise the discounted present value of profits over a finite planning period. Within the model decision making is recognised as a joint selection of rates of investment and resource input levels which together determine harvest output. The formal mathematical representation of the decision making is in the form of dynamic programming requiring appropriate algorithms to find the optimum solution. Whilst other social and institutional factors are recognized as relevant they are rarely incorporated into formal models unless they can be represented in some simple constraint or cost form. Heisay et


173

al. (1993) recognises that decision making usually involves the possibility of transferring from one crop mix to another due to changes in technology or prices for instance and that a key ingredient in the decision making is information provision and diffusion. In a study of the adoption of new high yielding wheats in Pakistan he identified farm-to-farm information diffusion, farmer’s age and literacy as major determinants of farmers awareness of new opportunities. Farm size also seemed to influence farmers’ willingness to experiment and hence the speed of diversification. A critical component in all models of crop choice is the representation of production possibilities (or more accurately farmers’ perception of these possibilities). Howitt (1995) provides a recent discussion that emphasises the need to recognise the complexity and subtlety of the relationships determining crop yield as a function of resource inputs, climate and topography. He criticises simple Leontief type models which relate resource inputs to harvest output in a linear and proportional form. These models, he argues, lead to large stepwise responses to changes in the price of inputs for instance, whereas we know that there is a gradual response involving changes in both input proportions and crop mix. He argues, therefore, for the adoption of more complex production functions such as Cobb-Douglas or CES developed by economists in other spheres of production economics. The complexity in any description of production possibilities has long been recognised. Heady (1952) describes some of the influences of product complementarity on harvest yields and Pannell (1987) explores some of the benefits found from crop and livestock rotation. He also draws attention to the fact that the pattern of resource use (particularly labour and management) has a strong seasonal pattern which may influence resource inputs and harvest yields. Decision making concerned with perennial crops can be more complex than with annual crops. Returns from perennials such as fruit may take a number of years to emerge after first planting; costs and revenues have to be seen in profile over a number of years forcing the model to take into account how farmers discount future money flows with all the complications of time horizons, inflation and uncertainty. French (1962, 1971) examines lemons as a perennial crop with separate equations determining farmers decisions to plant and to remove. Nerlove (1989) concentrates on annual crops in his detailed study of response to price changes. Some models focus on one crop only and its response to prices, yields etc. rather than the mix of crops in a region. Others (Bauer, 1990, Horner 1992, House 1987 for instance) provide examples of regional models of the USA, Canada and Turkey which predict the mix of crops over a large area. These regional models gross up in some way the micro decisions of individual farmers. An important component in a number of models is uncertainty. Hassan and Hallam (1990) draw attention to the fact that uncertainty exists in money returns due to uncertainty in expected market prices at harvest time and uncertainty in climatic conditions which leads to uncertainty in yields. Dorfman and Heien (1990) provide some examples of the level of production uncertainty and adjustment costs in the American almond industry. Anderson (1974) and Antle (1983) also point out that crop yields are stochastic in a region and simple averaging may mask important diversification of yields within the region and even at farm level. Just and Silberman (1986) provide an analytical framework to handle uncertainty of net returns. Using portfolio theory they show how it can be used to explain a preference among risk averse producers for product diversification. They also show that whilst diversity in crops may reduce the variability of profits it also reduces expected net returns (or net profits). One type of model draws a distinction between the deterministic and stochastic component in decision making. Using a probit or logit formulation these models assume some mean value for

174


Figure 10–4 An overview of the crop choice model.

certain factors (for instance the crop price vector, input prices and production function for the different crops) and this is known as the deterministic component. Other factors (for instance the uncertainty of yields, the variation between farmers and between location within an area) are represented as a statistical distribution and referred to as the stochastic component. The form of these functions make them amenable as allocation models; allocating areas to different crops based on the relative attractiveness of their deterministic component. Kraker and Paddock (1985) provide an example using the logit model to explain arable crop choice in Canada. A slightly different example is provided by Bewley et al. (1987) who use a logit model to explain the change in the share of wheat, barley and oats in the UK over time. The share of a crop in any particular year is a function of returns and rainfall of the previous year. Perhaps the most sophisticated model is provided by Moore (1994). He adopts a probit based model of crop choice and land allocation between five crops for the South Western United States. The model also incorporates the simultaneous choice of crop and irrigated water demand that is itself a function of the price of irrigated water. The Form of Crop Choice Model Adopted The Crop Choice model that is adopted for the Argolid contains four key relationships (Figure 10–4). 1. Crop Production. This relationship describes the production possibilities or production achievable in different areas for different crops taking into account the topographical and climatological conditions in each area. 2. Crop Cost. This relationship calculates the overall cost of production taking account of the input requirements and the price of these inputs.


175

3. Crop Profitability. A relationship that compares the market price with the cost of producing different crops in different areas 4. Crop Choice. A relationship which describes the choice of crop and land allocation in different areas as a function of profitability. The model generates the area devoted to different crops which then determines water requirements. At the same time crop choice is influenced by water availability and quality variables. The model contains a deterministic component which includes figures on factor costs, crop prices, linear input requirements and decisions based upon profitability. However it also includes a stochastic element which recognises that other factors also contribute to the final decision on crop choice. Crop Production Rate (Annual, Predicted, Units—kg/ha) The predicted rate of production for a particular crop in a particular zone depends upon the land quality in that zone. Land quality includes all those factors in a zone that may influence the productivity of a crop. It includes the type of soil and its component nutrients in the zone, the soil’s water holding capacity (which in turn depends on soil characteristics such as slope, depth, texture and nutrient content) and aspects of topology and climate. Each zone is therefore allocated a land quality index (the highest quality is given an index of 1). The reduction in output for a particular crop is then a function of this index. Poor quality land will affect the production rates of different crops to a different extent. The factors are derived from an FAO publication (FAO, 1985) supported by local knowledge. In addition to these factors concerned with the land two factors that may reduce production are included, a shortfall in the water needs of the crop (due to rationing or high cost) and water salinity. These factors are combined in the equation shown below. Predicted production for the relevant time period (in this case one year), is assumed to be a function of predicted shortfall in water and salinity—land quality is assumed to be known. Production is defined in units of kg per hectare—the issues of quality and time of cropping, which for most crops affect the price received, are not considered.

where subscripts: i=crop type; j=zone, k=land quality and =predicted production rate per year for crop i in zone j—kg/ha Exogenous Exogenous pi=production rate for crop i with highest land quality (=1) fik=output reduction for land quality k crop i Exogenous gjk=variable to represent land quality in zone j (1 for land quality) Exogenous =production reduction for salinity, crop i in zone j see below =production reduction for water shortfall, crop i in zone j see below Actual production in a time period is based upon actual salinity and actual water shortfall

176


Salinity Function The salinity function represents the reduction in production as salinity increases. It varies by crop and is strongly non-linear (see for instance Umali, 1994). Predicted production reduction is therefore a based on the salinity in a zone in the previous time period and a crop specific salinity function viz. where sjj(t)=salinity in zone j (µmho/cm) in year t Salinity is derived from the water model. Considerable evidence is available on the effects of salinity on production for different crops. Water Shortfall Function Similarly the water shortfall function represents the reduction in production with the shortfall in ideal water consumption, which also varies by crop and at certain levels is also non-linear. Predicted production reduction is therefore based on the water shortfall in a zone in the previous time period and the water shortfall function viz. where kj(t –1)= proportion of water requirements consumed in previous time perio water shortfall is derived from the water model. Considerable evidence is available on the effects of water shortage on production for different crops (see for instance FAO, 1987). Crop Cost (Annual, Predicted, $/ha) Crop cost is defined in terms of cost per hectare for different crops. It is assumed to be independent of land quality and actual production. The critical variable is the area of crop cultivated and the water consumed. Predicted total cost ($/ha) is therefore split into a basic cost (representing sowing, pruning, fertilisers, picking etc.) and a water cost. The equation is

where Exogenous ci=basic cost of crop i—$/ha =predicted water cost of type l at zone j—$/l =predicted proportion water type l used at zone j Exogenous di=water requirement crop i—l/ha Predicted water cost and proportion of water type used are derived from the water model and assumed to be as the previous time period. Crop Profitability (Annual, Predicted, $/ha) The predicted profitabiliy of a particular crop i on a hectare of land in zone j is given by the difference between the revenue and cost viz.


where =predicted profit rate crop i in zone j—$/ha =predicted production rate crop i in zone j—kg/ha ri =revenue from crop i—$/kg =predicted cost of crop i in zone j—$/ha

177

from above Exogenous from above

Crop Choice—Area of Crop (h) The area of crop devoted to a particular crop in a zone is assumed to be a function of its profitability relative to other crops. However given the large number of factors that may contribute to profitability and the variation that may be found in some of the variables listed above profit is assumed to contain a deterministic element (the derivation of which is described in the above equations) and a stochastic component containing all other elements. Assuming that this stochastic component is approximately Normally distributed (strictly Weibull distributed) then the share of different crops in an area is given by

where hij=area of crop type i in zone j—ha Exogenous hj=total area in zone j—ha from above =predicted profit rate crop i in zone j—$/ha β=empirical coefficient The second term on the right hand side defines the share of each crop in a zone which is multiplied by the zone area (hj) to give area of each crop. The coefficient β reflects the relative importance of the stochastic component and can be estimated by a comparison of observed crop shares and observed profit rates. If β is small this corresponds to a large stochastic element and crop shares are relatively insensitive to relative profits Whereas if β is large then shares are sensitive to profits and in the extreme case the crop with the highest profit is the only crop in the zone. Water Requirement (Monthly, l/Zone) A number of studies have examined water use of different crops in the environment of the United States. They also provide evidence on water use adjustments in response to changes in prices and changes in entitlement. Nieswiadomy (1988) and Ogg and Gollehon (1989) identified the price elasticity of demand for irrigated. Moore and Negri (1992) examined the impact on water use of a reduced water entitlement. Changes in the price and availability of water can affect the development of irrigation and the irrigation technology adopted. A number of studies (Caswell and Zilberman (1985), Negri and Brooks (1985), Schaible et al. (1991) have identified these relationships from empirical data. A recent study (Moore, 1994) found in an econometric analysis based in the United States that producers’ response to an increase in water price was ‘at the extensive margin (cropchoice and land allocation decisions) rather than the intensive margin (short run water use decisions)’ (pp. 872). The assumption used in this model is a linear water requirements function that

178


is not sensitive to price but is affected by any shortfall in availability. Crop choice is affected by water cost through the total crop cost calculation (the extensive margin). Total water requirement per zone is derived from the area devoted to each crop in that zone multiplied by the respective water requirement of that crop.

where vj=total water requirement—1 di=water requirement l/h The water requirement may be met from various sources including rainfall, springs, aquifers and canals. A shortfall may occur due to shortages or unavailability. Empirical Estimates 1991 The crop choice model described above requires a number of input variables, two functions and a calibration (of β) before it is operational. There are seven exogenous variables referring to area in zones (hi), land quality (fik gjk), potential crop production (pi), water requirements (di), basic crop cost (ci) and price (ri). Other parts of the Crop Choice Hydrology model provide estimates of water consumption (kj ulj), cost (wlj) and salinity (sj). Two functions (ujvj) allow estimates of production loss due to salinity and water shortfall. The output of the model is profit (rij) and production (pij) for each crop in each zone and the associated crop area (hij). Crop Shares in 1991 Details on the total amount of land devoted to different crops is available from official government sources and was collected for 1978, 1984 and 1991. Figure 10–5 describes the cropped area of twenty seven administrative regions in the Argolid for 1991 and the change since 1978. There is a considerable difference in the areas (defined in stremma which is 0.1 hectares) but little change over the period of thirteen years. The change in area under crops since 1978 is shown in Figure 10–5 and the share of different crops for twenty seven of the administrative regions of the Argolid in 1991 is shown in Figure 10–6. The crops are aggregated into four groups—fruit trees (of which oranges account for about 95%), olives, cereals and vegetables. The last group contains a large number of different crops with tomatoes being the largest. In the figure the areas are ordered according to share of fruit trees. It can be seen that in ten areas (Pyrgela to Drebano) fruit trees account for more than 80% of the area cropped. In seven areas (Koutsopodi to Fychtia excluding Nea Kios) olives account for more than 40% of cropped area. In only five areas (Kefalari to Skafidaki plus Elliniko do cereals account for more than 30% or more of the cropped area, and only Nea Kios has more than 15% vegetables. The reduction of the twenty seven villages to seven zones is described in Chapter six. Areas with very high fruit tree shares were incorporated in the same zone. Whilst there was some dilution of the variation in crop shares found over the region (Figure 10–6) this was limited and the seven zones used for modelling display distinct patterns in their crop shares (Figure 10–7). These crop shares are the main variable that the crop choice model aims to predict.


Figure 10–5 Area under crops in 1991 and change from 1978.

179

180


Figure 10–6 Share of crops by area (1991).

Input Variables Area in zones (hi),

The total cropped area is 21477 h. The seven zones vary in size from 2214 h (zone 4) to 3667 h (zone 2). Land quality (fik gjk)

The definition of land quality was derived from a number of sources. In particular it took account of soil type, irrigability of the soil, altitude and topology. As was shown in the discussion on zoning


181

procedure, confirmation of the derived quality index was provided by an agronomist based in the Argolid. The index varied from 1 in zones 1 and 2 through 0.95 (zones 5 and 6) to the poorest 0.67 in zone 4. The land quality index then forms the basis for the reduction in production likely to be experienced in different zones. Production of fruit trees would fall approximately in line with this index whereas production of olives would be affected less and vegetables most seriously. Potential crop production (pi)

Potential crop production is defined as that production that is achieved in the zones with the highest land quality index. These figures are available from the survey of 203 farmers with supporting evidence from an agronomist. Production loss due to salinity (uj) and water shortfall (vj)

The estimate of water shortfall (kj) and salinity (sj) is derived from the hydrological section of the modelling.

Figure 10–7 Actual crop shares in the seven zones.

182


Figure 10–8 Modelled crop shares in the seven zones.

Figure 10–9 Difference between the modelled and actual share in the seven zones. Water requirements (di)

Water requirements are derived from the survey of 203 farmers. In the model only fruit trees and vegetables are assumed to require water from other than rainfall (In practice a small proportion of these crops—less than 20%—receive small amounts of water from other sources). The total water requirement of fruit trees is assumed to be 1100mm on each cropped area distributed over the months May to September. A small part of this requirement may be met by rainfall. Vegetable


183

requirement is assumed to be 2200mm over the cropped area. This is spread evenly throughout the year. Basic crop cost (ci)

This is estimated from evidence gained from farmer interviews and advice from agronomists. Price (ri)

Evidence of the price received was obtained from the survey of 203 farmers which was reported in Chapter four. A wide variation was recorded, even within zones, and this reflected differences in the quality of produce and picking times and the fact that different markets attracted a range of prices. Water cost

Estimates of water cost (wlj) taking into account different types of water used (ulj) are derived from the hydrological sections of the model. Estimates are also obtained from the survey of farmers. Crop Choice Model Output The variables and relationships outlined above allow a derivation of production (pij) and profit (rij) for each crop in each zone. Using this information the parameter β can be calibrated. The value chosen (in this case – 0.06) is the one that gives the closest match of predicted crop share and actual crop shares. Figure 10–7 and Figure 10–8 show the actual and predicted crop shares for comparison and indicates the difference between the two. It can be seen that whilst the model reproduces certain aspects of the crop share pattern (in particular the relative shares of the four crops) the pattern between zones is not always reproduced accurately. Generally the high level of specialisation in zones is not always reproduced. The crop choice model only forms part of the overall model for the Argolid. Crop choice influences water consumption and at the same time water availability and cost influences crop choice. The strategic model of water flow, which is the subject of the next chapters, describes how the demand for water, and its supply, are related in the Argolid. This will take into account such factors as climate, acquifer performance (including seawater ingress) and man-made water distribution.

184

11. TOWARDS A STRATEGIC COMPLEX SYSTEMS MODEL OF THE WATER/SALT SYSTEM Roger Seaton

INTRODUCTION The first part of this chapter indicates that many of the technology focused scenarios depend upon an understanding of the history of the water/salt system in the Argolid. A distinction is made, therefore, between a water/salt model appropriate to the strategic explorations of a complex systems model and a hydrological model based on detailed measurement taken over many years. In the following sections a water/salt model is developed which is compatible with the seven zone strategic model. It draws upon data from the survey of farmers in the area and aggregate data from the Agricultural University of Athens and other sources (i.e. meteorological service). What this has to do, alongside the model of farmers’ crop decisions (chapter ten), is to reproduce the water/salt history of the region. The analysis follows a sequence as follows: • Stage 1 Analysis of bore-hole and salt data at the overall regional level. • Stage 2 Analysis of bore-hole and salt data at the village level. • Stage 3 Formation of the seven zones for the complex systems model and the production of zone histories. • Stage 4 Collation of aggregate data on water and salt at the regional level for “calibration”. All of the analysis in this chapter was undertaken on spreadsheets and with the use of Statistical and Chart Tools. Regression analysis has been extensively used and it is important to make some comments about the technique at this stage. The main effort in this chapter is to provide a quantitative history of physical change in water and salt as a basis for informing the strategic complex systems model. Such interactive models are not calibrated in the fashion of trend extrapolation models and depend for verification on their ability to reproduce a “complex history”. It was anticipated early on in the research that such a strategic model could be best structured on between five and ten zones (see chapter six). The following analysis both identifies significant changes in the main physical variables and contributes towards the selection of a set of zones by identifying the variation in behaviour of those variables in different locations. This enables the definition of an interface between such a strategic model and a detailed hydrological model. Precise engineering detail is not required for such activities, nor is it supportable from the data. Therefore, best fit equations have been produced as part of the process, sometimes for very limited data. Given these objectives and limitations, no attempt has been made to pursue overly sophisticated analyses.

186


Where more advanced techniques may be appropriate in further work on the Argolid or for more comprehensive data in other locations an attempt has been made to note them. DATA STRUCTURE The background to the farmer survey and the questions used for it have already been presented in chapter four. The interviewers were able to obtain information on virtually all bore-holes owned or used by farmer respondents and to identify when they were first drilled, their depth, the flow they supplied when opened, and whether and when they had failed and why. Farmers also generally had good factual knowledge about the salt content in their wells and bore-holes. Basic data was collected about some 620 wells and bore-holes on respondents farms. Additional information was also collected on more than 200 bore-holes which are associated with land on which the farmers’ main crop is grown. The data was held on two master files which were edited for specific analyses within crop decision making and water/salt models. A typical extract from the data files, for the first stage of the analysis presented in this chapter, is given below (Table 11–1) with a brief outline of the coding classification which was adopted. Interview number: Each farmer was allocated an interview number which facilitates the linking together of the various data files between the different uses in the project. Village: The villages in the region were numbered and the second column refers to the home (village) location of the respondent. Year opened: The year in which each bore-hole or well was drilled was recorded. Depth metres: The depth at the time of opening was also recorded. In a few cases a specific borehole had been deepened but this was so infrequent in this data set that the original depth has been used. Account will be taken later of the altitude, above sea level, of the top of the bore-hole. In the first stages of analysis the focus is on change and rate of change of bore-hole properties. Initial flow cu.mts/hr: Farmers are very alert to the properties of their water sources and were able to give the maximum flows from bore-holes when they were first used. In some cases the maximum flow may be limited by the capacity of the pumps; in others, and with increasing frequency in the early 1990’s, bore-holes failed to yield water at the outset (Figure 11–2). Table 11–1 Sample data from first stage analysis.

TOWARDS A STRATEGIC COMPLEX SYSTEMS

187

Current flow cu.mts/hr: Similarly the existing flow is usually known and in many cases there has been a reduction sometimes to zero. Where a bore-hole flow had at some earlier time ceased to flow but the year in which this had occurred was not known the code 888 was used. 888 was generally adoped to represent missing variables in the data set. Failure type: For simplicity in this presentation the raw data has been condensed to a simple code. Many of the respondents were able to give the current sodium chloride content of bore-holes, which have water flow, in parts per million (ppm). This is not the same as recording total salt content but offers a useful and consistent measure throughout this chapter. Where the salt content had precluded its use for agriculture this was also noted for all bore-holes, although the year in which the water ceased to be used was not recorded. The coding for bore-hole failure was as follows: 1. no change in flow 2. a reduction in flow but not to zero 3. reduction in flow to zero 4. no flow when drilled and no flow now 5. too salty to use Village of main crop: In the initial exploration of the data it was not necessary to differentiate between villages. However when some spatial analysis was required it had to be in terms of the location of the main crop and the properties of the water that was used to irrigate it. Further investigation showed that most farmers owned or farmed more than one plot of land and that many of these were not contiguous. Therefore a proportion of plots for the main crop were in villages other than the one in which the respondent lived, further analysis uses that location data. This difference in location led to a small error in early analyses and will be returned to later in the chapter. Stage 1: Regional Properties The main point of interest in the early analyses was the extent to which water had become more difficult to obtain in the region (Figure 11–1). This shows the change in depth of bore-holes over time differentiated by those which provided flow when drilled and those which were dry when drilled. The most striking feature is the marked change in depth over time. Before the early 1960’s the trend towards increased depth was considerably less than later. This was the period during which the natural recharge more or less balanced the extraction of water. Most of the extraction up to 1950 was from conventional wide bore wells (some of which were horse drawn) in the central plain and in areas close to the sea. This data also includes a few villages to the valley to the East and outside the main region which were known to have suffered from salt intrusion into aquifers many years ago. After this period there is a steep increase in depth and a marked increase in the number of bore-holes which were dry when drilled although in 1994 there remained a proportion of successful and shallower bore-holes and a broader spread of depths even for those which are dry upon drilling. The more recent data suggested the influence of a number of prevailing factors in 1994.

188


Figure 11–1 Depth and success of wells/bore-holes when initially drilled.

– A trend for irrigated agriculture to spread from the flat and low central plain into the higher peripheral areas. Hence the subsequent need to analyse the data spatially, and by altitude. – A reduction in the level of some aquifers with respect to sea level and the increased failure, at a given depth, of finding water in some areas particularly towards the periphery of the region. In some of these areas farmers may have given up trying to find water. – There was a suggestion that farmers in other areas, particularly those at lower altitudes but further from the sea, were able to drill relatively shallow and successful bore-holes. It is remarkable that bore-holes in excess of 400 metres have been drilled. Not only is the cost of drilling high (and needs quite sophisticated American drilling technology) but the cost of pumping from such bore-holes is also significant. This data therefore could support further analyses to correlate cost and energy functions with depth and success. Figure 11–2 concentrates on those bore-holes which have failed in some way. Those with a reduction in flow are too few to be worth plotting as a separate group although the graph incorporates the “dry when drilled” bore-hole data identified in the first chart. The main trend here is the very early evidence of bore-holes which have become too salty to use since they were first drilled. These “too salty to use” observations are the result of sea water intrusion into the aquifers of the main plain. As might be expected in a situation of progressive searching for water from aquifers that are becoming lower, the overall trend in unsuccessful bore-holes is generally at a greater depth than for those which have ceased to provide water since drilling. Stage 2: Village Analysis All villages: A village by village analysis was undertaken of the date when bore-holes were opened and the depth to which they were drilled. This was based upon the data collected from interviews


189

Figure 11–2 Failures in wells/bore-holes.

Figure 11–3 Regression lines by depth of bore-hole and date first used—for all villages.

and is represented in Figure 11–3 as regression lines for each village. While it is accepted that the figure is of little value for examining individual villages it does provide an overview which highlights a number of interesting features. These are: • The enormous variation in the gradients of the lines revealing large differences in the rate of change of depth at which farmers have attempted to find water over time in different locations. Although some of this variation could be due to systematic variations of the land altitude of boreholes over time within villages, the variations of the steepest lines are outside the range of altitudes measured on contour maps of the area. This is also largely true of the shallower lines

190


which are associated with villages in more or less flat areas in the central plain close to the sea. In addition the range of depths is substantial between villages. For one village to show changes from 20 to 300 metres over 30 years and another to show change of 10 to 30 metres over 70 years suggests some interesting physical and spatial dynamics. • Many of the steeper gradients have starting dates later than those with lower gradients and a number of them end at earlier dates than the majority. In part this may be due to “sampling” limitations, and the survey did not contact a comprehensive enough range of farmers in certain communities. It does, however, suggest the proposition that there is a time dependent spread of irrigation from bore-holes in lower altitude villages to higher altitude villages. In the Argolid this would correspond to a spread from the inner plain to the outer plain and thence to the peripheral higher altitude and hillier areas. Failure of bore-holes: In this sequence the incidence of the various types of bore-hole failure is examined as part of the process towards establishing the seven spatial zones adopted for the complex systems model. The data format is shown in Table 11–2. In this data set which uses the example of Argoliko, a simple approach has been taken in the early stages of failure analysis. The failure types identified above are adopted here and simplified further into a code of 0 for no failure and 1 for any type of failure. For this example there are a total of sixteen observations and the “Cumulative failure” has been divided for each successive year by the total number of observations giving a “Contribution to total failures”. Table 11–2 Cumulative failures as a proportion of total observations (Argoliko).

It should be noted that the data is based on two pieces of fundamental information: what the bore-hole was like when opened and what it is like now. The precise date on when a failure developed depends on the nature of the failure. A bore-hole which was dry when drilled can only be dated as accurately as the data offered by the respondent. However a bore-hole which has reduced flow, become dry or too salty could have done so at any time within the period between drilling and


191

now. No systematic attempt was made to encourage farmers to be more specific during the interviews. In view of this limitation a sequence of analyses was undertaken of the separate failure types at this and the final stage of the investigation. Figure 11–4 also presents an overview of the villages rather than being concerned with the specific characteristics of each community. The first and last points on the “Contribution to total failure” have been taken and connected with a straight line to give a simplified presentation of the data. The x-axis has a number of interesting features which are followed up later. For instance, the range across the villages of the total proportion of all bore-holes which had failed by the end of the recorded history of each village varies from 0 to 1. There is also considerable variation in the time periods over which the failures took place with some villages starting and finishing their failure history over very different durations. This reflects early bore-holes which became salty, those that ran out of water and more recently the difficulty of finding water at all in some of the outer villages. The next stage of analysis was to examine the characteristics of the bore-holes spatially by linking the data to the seven zones described in chapter six. Stage 3: Analysis by Zones There are some differences in actual zone sizes and shapes and the length of boundaries between zones. In particular, this may be misleading where the zone borders the sea and is vulnerable to sea water ingress. As will be seen there are sufficient spatial differences in the water/salt history within the zone structure to enable further analysis in support of the complex systems model. Zone 1 is the

Figure 11–4 Simplified chart of failures as a proportion of total bore-holes.

most internally disparate zone containing land near the sea at low altitude with quite high ground further inland. This results in more noise in some of the analyses for this zone than for the others. As will be seen in the complex systems sub-model of the water/salt system, this is not critical since the zone is relatively separate geologically from the other contiguous zones and therefore does not interact substantially with them. It is also the zone which has had the least problems with water loss

192


and salt since its aquifers are recharged from a different source, namely the wetter Tripolis mountains to the west.

Figure 11–5 Dry bore-holes when drilled.

Figure 11–6 Bore-holes with flow when drilled but dry in 1993.

Depth and failure by zone: The data used in the analysis for Stage 2 was collapsed into the seven zones for the period 1963 to 1993. This gave 513 items of valid data on separate bore-holes and is shown in Figures 11–5 to 11–8. Each figure shows regression lines for bore-hole depth as a function of year for those zones about which there is sufficient data. The gradients of the lines correspond to change in depth in metres per year. Given the number of observations and relatively “noisy” data the R squared values are not in themselves important, however, they constitute the only viable way in this data set of deriving an equation suitable for use in subsequent modelling. Nor has it been worthwhile attempting to fit non-linear functions. In all of these charts the absolute depth is not itself relevant since the zones have considerable variation in altitude and hence the height of bore-holes.


193

Figure 11–7 Bore-holes with unchanged flow.

Figure 11–8 All bore-holes.

Figure 11–5 shows that Zones 1 and 5 have the least rate of change in depth associated with this particular class of bore-holes and Zones 3, 7 and 4 the greatest. This reflects the point made earlier, that Zone 1 has relatively few problems while Zone 5 is one of those in which water is available but parts of which were affected by salt early on in the thirty year period. The reduction in depth in Zones 3, 7 and 4 is attributable to the loss of aquifers in the peripheral foothills which, ironically, provide the majority of the natural recharge for the region. The rate of change in Zone 4 is the greatest followed by Zone 7 but both cease to have recorded data just before 1990. The suggestion from the field work is that the pursuit of water from aquifers has been given up in Zones 4 and 7. Zone 3, however, has a mix of territory; some is really part of the central plain in Zones 2 and 5, while the majority is identifiable with the peripheral hills.

194


Figure 11–6 again shows that Zone 1 has suffered no reduction in the flow of bore-holes whenever drilled and Zone 3 is again the zone which has most severely lost access to water in its aquifers. Despite the lack of information on Zone 4 which is due to the small number of observations the data suggests that aquifers have dropped by almost 180 metres in the period from 1970 to 1988. Zones 2 and 6 again manifest no apparent problem of water depletion although in Zone 6 there is a problem of quality. In Figure 11–7 Zone 1 shows little change in depth over time for bore-holes that have remained steady in their rate of flow. The interesting zone in this figure is Zone 6 which has the greatest rate of change in flow as well as in depth. However, this is not because of one single lower aquifer but because this zone and part of Zone 5 sit above a complicated set of aquifers. Those bore-holes nearer the surface were contaminated with sea water in some areas of Zones 5 and 6 as far back as the 1960’s. Some farmers in these zones sought clean water by drilling deeper, however wherever possible they tended to use water from the Anavalos infrastructure. Some parts of Zone 5 were less vulnerable to salination and have been able to tap into water from successively deeper bore-holes. The final chart in this sequence is for “All bore-holes”. Zone 3 had the deepest bore-holes but largely because the land is at a greater altitude. The greatest rate of change was for Zone 4 (loss of single aquifer), then Zone 7 and Zone 3 for largely the same reason and to a lesser extent Zone 2. Zones 5 and 6 reflected the problems with quality rather than quantity. Zone 1 had the lowest rate of change of the peripheral zones and this rate was probably as much due to the increase in irrigated agricultural land in the higher areas as with increased depth of bore-holes in any one location. Stage 4: Regional Aquifer Analysis Altitude: Changes in the depth required to access an aquifer is an important symptom of the deterioration of the water/salt system that irrigated agriculture depends upon. However, at some stage in the understanding of this behaviour, and its interactions with the other systems, there is a need to comprehend wider systemic processes. While analysis of the relative differences in rates of change would be helpful, ultimately the region can be seen as a stock of depleting water into which sea water has ingressed over many years. It is necessary to estimate the absolute changes in the aquifer volumes in order to approximate the volume of sea water ingress, and therefore, changes in the salinity of various parts of the aquifer system. The relative altitude of land from which boreholes are drilled is therefore relevant in as much as there are interfaces between the zones across which not only clean water flows but also that which has been contaminated by sea water. The absolute height of aquifers above sea level needs to be known as part of this calculation and there Table 11–3 Average altitude of zones.


195

Table 11–4 Salt content in ppm by zone (1994).

are two ways of addressing this. The first approach was to examine in detail a contour map of the region and the territory of the villages and to estimate the low/medium/high altitudes of the land. This was undertaken as part of the zoning procedure although the values obtained appeared to be consistently too high. The second approach was to examine the depth of the earliest bore-holes. These would be drilled to the minimum depth at a time when aquifers were highest. A variety of calculations were undertaken on data prior to 1963 and then on all the data. Regression equations proved unhelpful due to poor correlation coefficients. In any event there are sufficient differences between villages to support the need for a more detailed approach. Using the estimated values of regressions every value of depth was examined for each village and adjusted as required to avoid negative or very small values. As a check on this somewhat arbitrary procedure a regression analysis was undertaken between the “average zone altitudes” (Table 11–3) derived from this approach and the average of those obtained from maps. A very good fit was obtained which shows that the second approach consistently gives values equating to about 55% of the map data. These values are simply subtracted from the constant in the equations used in subsequent modelling. Salt: Many farmers were able to provide figures about the bore-holes that were being used to irrigate their main crop.1 The data from the survey has been collated by zone and is shown in Table 11–4. It can be seen that zones 2, 5 and 6 had averages above the level of damage to crops (i.e. greater than about 450 ppm.) with maximum values reaching 50% of sea water chloride content (typically 9000 ppm). From the data (Table 11–4) and comments in interviews a very crude estimate can be made of the changes in salt content for the zones from 1963 to 1993. This is shown in schematic form in Figure 11–9. The impact of salt on Zone 5 relative to Zone 6 was probably due to geological access from the sea and to the early abandonment of irrigation bore-holes in Zone 6. Also the increase represented in Zone 7 is based on very limited evidence and as such must be treated with caution. One other feature which the complex systems model needs to capture is much more obvious. The peripheral Zones 1, 3, 4 and to some extent Zone 7 are not experiencing salination. In Zone 1 this is because it has an adequate supply of good quality water and is geologically separate. In the other Zones it is because the natural recharge aquifer has dropped in level. Recent anecdotal information, however, suggests that salt has been found in very deep (>400m) bore-holes in Zone 4 and that it may not be from the Argolid area but from the north towards the coastal area near Corinth.

1

Information was only obtained about the main crops of respondents.

196


Figure 11–9 Estimated changes in the salt content of aquifers by zone. Table 11–5 Salt content of water from the Anavalos canal.

Anavalos: The water from this source is currently being used at the rate of twenty five million cubic metres per year and this will increase substantially with ongoing construction. As has already been discussed, this water also has a variable salt content as indicated in Table 11–5. The overall trend has been upwards although the lower value in 1989 was due to a wetter winter and the higher one in 1992 to a dryer winter. Salt levels have again decreased considerably in 1996– 97 with the high rainfall however, while this variability is of concern so is the fact that the water can reach salt content levels which, if used directly for irrigation, could cause as much as a 30% reduction in the yield of citrus trees. Aquifer volumes: The data from farmers provides a indication of the change in aquifer depth in the zones and rates at which water was extracted. This can provide an estimate of the total decrease in volume of water in aquifers, however, without some knowledge of the original volume it is not possible to estimate the volume of sea water ingress that has led to salination. The complex systems sub-model of water and salt needs to be able to show how different types of intervention have or would affect this crucial factor. Measurements taken over many years by the Agricultural University of Athens (AUA) have led them to make the following estimates:


Original capacity of aquifers Loss over 30+years Aquifer use in 1992 Use of spring water in 1992 Natural recharge

197

2×109 m3 1×109 m 95×106m3 per year 25×106m3 per year 50×106m3 per year (last in balance in 1960).

This data can be compared with that collected from farmers for this study to suggest that water drawn from the aquifers over the survey area was 120×106m3 in 1993. It will now be seen how the water and salt data presented in this chapter was incorporated into a strategic model of water flow.

198

12. THE STRATEGIC MODEL OF WATER FLOW Peter Allen

INTRODUCTION The previous chapters have interpreted the water and salt profile for the Argolid Valley and have considered how this, in conjunction with other attributes, has affected the choice of crops grown. The next two chapters will introduce ongoing work into the strategic modelling of water flow and the progress towards a dynamic representation of the complex relationship between the natural and anthropogenic components of agricultural production. As was explained in the preface some of the mathematical representations may prove unsettling to the casual reader. However, it is hoped that by reading around these, and drawing upon the background knowledge of the Argolid situation, such a reader will gain an insight into the modelling process and a sympathy for the contribution that such representations can make in the exploration of possible futures. The flows of water and of salt across and through the Argolid are obviously three dimensional, and have been modelled by dividing the total horizontal area into the seven spatial zones, and the vertical dimension into three sections. The horizontal flows depend on the slopes and the geological structure, so that different zones are linked with widely varying connectivities. The flows of water through the area of study was then modelled by considering the 3-D movements of water onto and off the surface, through the subsurface layer if it is permeable, and in and out of the aquifer. Each zone was therefore viewed as having three vertical layers: a surface layer where the crops are grown, a subsurface layer which may or not be permeable, and an aquifer (Figure 12–1). The main anthropogenic driving force of water and salt flow arises from the decision to grow irrigated crops, and the need to pump water up from wells, bore-holes and from canals to maintain growth during the hot, dry summer. Otherwise, the winter rainfall has fed the aquifers and springs, and given rise to a net positive hydrostatic pressure throughout the area. The model considers the effects of pumping underground water as the area of irrigated crops has increased. Figure 12–1 shows the structure of the model with the flows to and from the surface as a result of infiltration and of pumping. It also shows the horizontal underground water movements that result from the differences in hydrostatic pressure between water in the different zones. The dynamic model that has been developed represents crop choice decisions as annual and sets out the agricultural requirements for water over the next twelve months. It is accepted that choices are not made in this way and that there is a marked temporal difference between those decisions relating to annuals and those which relate to tree crops. The annual decision not to change allows us to cater for this variation (i.e. by retaining tree crops). The amount of irrigation that will be required then depends on the profile of crop needs throughout the succeeding weeks of the year,

200


Figure 12–1 Representation of water flow between levels.

and the rainfall and the evapo-transpiration that actually occurs. The model therefore uses a short time step of three days (one tenth of a month) to describe the movement of water and salt over and through the different zones and sections of the model. It simply uses balance equations based on the water and salt accounts of each section. The crops that farmers choose have been divided into four types (see Chapter ten). What matters here is that two of them require irrigation, while the other two do not although more recently irrigated olives have become more prevalent in the area. The crops are: Fruit trees (Citrus, apricots etc.); Cereals, Vegetables and Olives. The fruit trees and the vegetables require irrigation, vegetables around 220 litre/kg per year of crop, and citrus 110 litre/kg of fruit. This means that the farmer will drill bore-holes proportionally to the hectares of crops one (fruit trees) and three (vegetables). In the dynamic model the rate at which farmers can switch between crops has to be considered. While they may set their aims at the start of each year, new plants can take over five years to reach maturity thereby slowing the rate of conversion. The model therefore incorporates a parameter to represent the rate of crop switching. We can now consider the water balance equations for the three layers of each zone. THE SURFACE LAYER The different terms that are quantified to calculate the changing volume of water present in the surface layer are as follows: – – – – –

rainfall run-in run-off non-crop evapo-transpiration crop consumption

THE STRATEGIC MODEL OF WATER FLOW

201

– infiltration – water pumped for irrigation – water pumped for frost control. Meteorological data has been used alongside the hydrological information provided by the Agricultural University of Athens to provide the average monthly rainfall and also the consistent geographic variabilities that are observed. In the model, short term variation is added using a random number generator. The index for increased or decreased rainfall in the different zones is shown in Figure 12–2. The equation used to generate rainfall is: where A(1, z) is the water available to zone z from rainfall, A0 is the annual rainfall and the remaining terms reflect the seasonality of the rainfall with small random variations for each timestep. The monthly temperature profile is used to calculate the rate of evapotranspiration, and the known monthly profile of water requirements of each type of crop are used to generate terms which transfer water from the surface layer into the atmosphere, and to some extent into the crop. Evapotranspiration, however, depends on the amount of surface water present, since clearly, if the surface were dry, none could occur. The term is: where Sw(z) is the volume of water in the surface layer of zone z.

Figure 12–2 The geographic coefficients of increased/decreased rainfall used in the model.

The water running into and out of the surface layer of a particular zone z, will depend on the rainfall, the slopes of the terrain, and the amount of surface water present in z as well as in neighbouring zones. The length of the border linking two zones is also taken into account. The rate

202


of run-off depends on Sw(z) the volume of surface water present in z. This term has been represented by: where sflow takes into account the morphology of the surface. The infiltration term depends on the amount of water in the surface layer, Sw(z), and the permeability of the soil. Figure 12–3 shows the pattern of permeabilities that have been used in the model and which have been deduced from a detailed study of the soil types. The term for the infiltration is therefore: The uptake of water by crops is calculated from their water needs per hectare per month. For each zone, therefore, once the ‘annual’ crop decision has been made, the crop demand per month is set, although in a wet year this may be met to a large extent by rainfall, while in a dry year it may require a great deal of irrigation. The term in the equation which describes the irrigation of crops simply reflects the fact that the needs of the crop are not being met by the amount of water in the surface layer. When the volume falls below some desired level, the pumps switch in and deliver irrigation water. Initially, this water came from shallow wells, but as the ground water level falls, these became bore-holes. The aquifers then furnished irrigation water for some time, until the incursion of salt water in specific areas led to the construction

Figure 12–3 The pattern of permeability per zone used in the model.

the construction of the Anavalos canal to bring water across the plain from the springs located at the south west corner. When the volume of surface water is above the desired trigger level, then no water is pumped for irrigation. However, when it falls below this level, the ground water is pumped at a rate that is proportional to the area of fruit trees and vegetables planted. In other words, it is assumed that the farmers who plant these crops, also install enough irrigation capacity to provide them with water in the summer. The cost of the water depends on depth gradually increases as the aquifer


203

reserves decrease with use. The exhaustion of an aquifer is represented by an escalation of costs to infinity. The rate of pumping when the surface volume falls below the desired level is given by: Where A(2, z) is the water available to zone z from pumping the aquifer, ha(1, z) are the hectares of crop 1 (fruit trees) and ha(3, z) the hectares of vegetables planted in zone z; pmp is the pump capacity required per hectare. As soon as the surface water volume return above the necessary level the pumps are stopped. In this way the model automatically takes into account the effects of an exceptionally wet or dry spell, and does not simply use the “average” rate of pumping as a fixed term. Thus, a particularly wet spring may delay the need for irrigation, whereas a dry one may advance it. The model can also take into account the choice of the “desired volume” of surface water which acts as a trigger, and naturally this will also affect the amount of irrigation that takes place. The integrated system will allow for the long term effects on economic yield of different choices of trigger levels, and this can be used to reflect the use of “flood” or “trickle” irrigation systems. As water levels fall in the aquifer, the incursion of sea water leads to an increasing salinity, and this is transferred to the surface with pumping. Crops have a different sensitivity to salt, but they all suffer a decline in yield and the continued use of salty water will result in the gradual destruction of the agricultural productivity of the land. When salinity begins to seriously affect a region, the farmers use their political voice to obtain supplies of fresh water. As has been seen this occurred in the mid-1960s for the coastal areas of the Argolid, and in recent years has led to the construction of a much larger canal to bring water from the very large natural spring of Anavalos to the areas suffering from lack of water and high salinity. The only difficulty with this plan is that Anavalos itself is fairly salty, and in recent years this has been increasing. If saline water is used to irrigate the plain as a whole then the salt will simply accumulate in the soils leading eventually to the total destruction of the agricultural potential of the region. Again, as has been described, recent heavy rain and the replenishment of aquifers with fresh water have reduced this problem for the present (1997). In addition to water being used to irrigate crops it is also used extensively against frost. As described in the earlier chapters, during winter, from November until March, water from the boreholes is sprayed into the air to try to raise the temperature of the air above freezing. This results in a considerable consumption of water from the aquifer even if the region is connected to a canal for its irrigation. This term is represented by:

(where FrA(2, z) is the aquifer water available to zone z for frost prevention), and is switched on using a trigger level of temperature with a random variation term. The complete equation for the change in the volume of water in the surface layer is:

where A(3, z) is the water available to zone z from the Anavalos canal system. The time step of the model is one tenth of a month (approximately three days) and the areas of each crop are revised annually for each zone, according to the perception of potential profits in the next year, and the rate of switching that is possible.

204


THE UNDERGROUND WATER This is divided into two parts, a porous layer and the aquifer. The porous layer has been treated very simply, with water infiltrating into it from above according to the soil permeability, and draining down to the aquifer at a rate which depends on the depth of the aquifer and the geology of the zone. The equation for water balance in the porous layer is therefore: where Pv(z) is the volume of water in the porous layer, dPv(z) the change, and Tau(z) the time to filter down in zone z. The aquifer water balance is a much more complicated equation. This takes into account the inflows of water from the catchment, the water filtering down from above, the water being pumped out by farmers, and the flows induced by the hydraulic pressures changing between the zones as water is pumped. The water running into and out of the aquifer is therefore dependent on the length of the border with the neighbouring zone or the outside, and also on the difference in heights of the water in each zone. The height in each zone changes over time as the volume changes, so that pumping water to the surface results in a drop in level and a consequent in-flow from neighbouring zones, or from the sea if the zone is at the coastal edge. The term describing the “run-in” of water is called: where border (z, w) is the length of the border between z and w, link(z, w) expresses the geological connection between the two, and H(z) the height (or hydraulic head) of water in zone z. The “runout” of water is the symmetrical term with w and z reversed. The other important terms are those describing the pumping of water from the aquifer to the surface, which are switched on when the surface soil is too dry and the surface water falls below the trigger value Swdes (the desired water volume). The aquifer equation therefore becomes: where daw(z) is the change in water volume of the aquifer of zone z, Pv(z) the volume of water in the porous layer of z, Tau the time of infiltration, A(2, z) and FrA(z) the water pumped from the aquifer z for irrigation or for frost prevention. This completes the water balance equations for the Argolid model. It shows how the levels in the different aquifers change as farmers irrigate their land from bore-holes, and how this in turn, leads to flows of water between the zones, and changes the cost of water as a result of increasing depths of drilling required, but most importantly it tells us how the sea will invade the underground water, and lead to salination of the system. SALT MOVEMENT EQUATIONS The medium which transports the salt around the system is water. Initially, before large scale irrigation occurred, there was a gentle, positive hydraulic head throughout the system, which meant that the aquifers were pure, and that there were some marshy areas of land. There was also a steady transfer of the catchment water to the sea. However, as the hectares of irrigated land were increased, the overall water balance of the ground water changed, and around 1960, became negative. Irrigation in the coastal zone rapidly led to the incursion of sea water into the ground


205

water, with a consequent transfer of salt. The continued pumping for irrigation transferred the salty water from the aquifer onto the surface, where, gradually the productivity of the soil was eroded. As was discussed in chapter seven, a first response was to build a canal to transport spring water from the western corner of the Argolid to irrigate the coastal strip. While this supported the continuation, and expansion, of intensive fruit tree growing the farmers further back from the coast also continued to expand the area of irrigated crops and hence the amount of water pumped from the aquifer. The lowered hydraulic pressures led water to feed back from the coastal aquifers, thus transporting sea salt underground some 10–15kms inland. Gradually then, the salt problem increased with farmers in the periphery finding it increasingly difficult to get water at all. There has therefore, been a demand for water to be delivered from the Anavalos system both to ameliorate salinisation, and satisfy water demand. The problem was, however, compounded by the increased salination of the Anavalos source. The equations for the movement of salt are therefore fundamental to the integrated model which must represent the transfer of salt between zones as well as that of water. In each zone, we then suppose that mixing and osmosis lead to a revaluation of the salt concentration so that:

In this way the model constantly updates the concentrations, and uses the values to determine the quantities of salt transferred with the water flows. For the surface layer, the salt initially present is the volume of water multiplied by the concentration. However, this water is diluted by rainfall and changed by the run-in from neighbouring zones, depending on the volume and concentration involved. The water pumped from the aquifer or taken from the canal will bring with it a quantity of salt depending on the amount of water put on the surface, and the salt concentration. Figure 12–4 shows the flows of salt show that the incursion of sea water into the aquifers leads to salinisation of the soils. The salt arrives on the surface either by being pumped up from the aquifer or with the runin from neighbouring zones. It leaves the surface of the zone either through infiltration or with the run-off. A critical factor therefore is the “retention” of salt in the surface by binding to soil particles, so that it leaches into the subsoil and stays there. Thus the change in surface salt is given by:

or more formally: The porous layer receives the salty water from the surface (infiltration), and passes it to some extent to the aquifer water (leaching). The term “retention” is significant because it refers to the fraction of the salt that is retained in the surface layer when infiltration takes place leading to the steady accumulation of salt in the soil. The only way to reduce this is for rainfall to leach out the salt either into the deep sub-soil or as “runoff”. In the zones of low infiltration, the situation is even worse, because all the salt delivered to the surface remains there, except that greater run-off may occur in the rainy season. The critical questions therefore concern the detailed mechanisms of water infiltration, salt retention and run-off. The transfer of salt between the aquifers of different zones depends on the flows of water between them. This is affected by the hydraulic pressure differences between the zones and this in its turn by

206


Figure 12–4 The incursion of sea water into the aquifers.

the amount of pumping and the inflows from the catchment. The vertical transfer of water from aquifer to surface results in horizontal water and salt transfers as a result of the differential pressures generated. Typically for example we have the equation describing the flows of water between zone six and its neighbours:

where A(2, 6) is the volume pumped to the surface in zone 6 FrA(2, 6) is the volume pumped to protect against frost damage border(6, z) is the length of the boundary between 6 and z link(6, z) is the geological connection between 6 and z ht(z) is the hydraulic head in zone z Pv(6) is the infiltration from the surface back to the aquifer Tau is the delay in this. From this we can calculate the flows of water between the neighbouring aquifers. This will be: If the hydraulic head is higher in z than in 6 then the flow will be from z to 6 and it will transport a quantity of salt: since salt(z) is the quantity of salt in aquifer z and aw(z) is the quantity of water. The ratio is therefore the concentration of salt and by multiplying this by the quantity of water that flows between z and 6 we find the quantity of salt that has been transferred. We assume that if ht(z) is greater than ht(6) then there is no salt transfer in the direction 6 to z. This assumes that salt transfer only takes place through bodily transport with water flows. It does not take into account the possibility of osmosis and of the transfer of salt through dilution into


207

regions where the hydraulic pressure may in fact be higher. A simple modification of the model could take this possibility into account if it was thought to be important. From the calculations of the quantities of salt transferred between each pair of aquifers we can calculate the total amount transferred to or from six.

where the amount of salt A(2, 6) * salt(6)/aw(6) is transferred to the surface by the irrigation pumping of the farmers. In a similar way the flows of salt through the system are calculated for each zone. As salt affects the growth potential of the different crops, however, the farmers can modify their decisions concerning the areas devoted to each crop, and in this way alter their irrigation requirements in response to the changing situation. This is an example of the operation of mutual feedbacks between the physical processes of salt and water transfer, the biological ones of plant growth, and the social and economic system inhabited by farmers. It is the integrated behaviour generated by this linked system that our complex systems model is trying to describe in order to support a strategic assessment of the implications of possible policies or changes in climatic or economic conditions—such as the increased rainfall and reduced price support that have occurred over the past three years. DYNAMIC RESPONSE OF FARMERS As described above (Chapter eleven), we can find a mathematical expression which indicates the number of hectares for each crop that farmers would like to plant and harvest, given the predicted prices, yields and costs of inputs.

The attractivity is of course a function of the profit that is expected, and this depends on the type of soil, the price of crops, the yield expected and the costs of inputs. Over time these change, with market conditions, subsidies, soil condition, salt deposition, and water availability, and as the relative attractivity changes so do the areas devoted to the different crops. But, of course, farmers cannot simply uproot their existing crops and succeed in harvesting another one immediately. Fruit trees, for example will take several years (>5) to mature and to reach reasonable levels of production, while olives will take longer. Although vegetables may give a fast return, they require a great deal more labour and uninterrupted irrigation, and so farmers must be sure that this will be available before increasing their committment to the crop. As a result, we have made the supposition that the area which changes each year is proportional to the difference that exists between the actual and desired land areas. So, for zone z and crop i, the change in hectares is: where hec(i, z) is the desired value at this time, and ha(i, z) is the actual size, al is a response rate for the adjustment process. The study reveals that the total hectares in each zone for agricultural use does not change very much over the last thirty years although there is a marked increase in the land that is devoted to irrigated crops. According to the soil types and climatic conditions of the different zones, there are differing advantages to growing the four crops represented in the model. However, the increasing cost

208


of getting water from ever deeper bore-holes and the higher levels of salt in the aquifers and surface water can lead to altered levels of potential profit and therefore to a changing allocation of land to crops. Another response concerns the ability of farmers to lobby for action if their crops are under threat. In the case of the Argolid, pumping for irrigation in the coastal zone led to the incursion of sea water, and to the salination of the aquifer and soil. Falling crop yields provoked a response in the form of a canal bringing fresh water from the springs in the south west of the plain and using this water for irrigation of the coastal zone. This proved very successful at maintaining production, and as the sea water and salt have penetrated ever further into the region, so these canals have been extended. This kind of farmer, and political, response has been taken into account in the model through the inclusion of a term which calls for the building or extension of the canal to a zone, when the following applies; aquifer water is above 400 ppm of salt and/or surface water salt peaks in the summer are over 750 ppm and starting to affect crop yields. The model then stops, asks if the “policy maker” running the model wishes to build the canal, and implements the decision that is taken. If it is affirmative, then on the screen a canal appears, and the farmers turn from pumping the aquifer for their irrigation to using the canal water. They still use the aquifer however, for spraying against frost in winter. If the policy maker does not build the canal, then at the end of each year the farmers repeat their demand, and in the mean time must of necessity respond to the situation by reducing their irrigated crop areas. A range of pricing policies for the canal water can also be represented and in this way the potential effects of these on the long term development of the region explored. In the case of Anavalos, as has been discussed elsewhere, there is an ample supply of water, but it seems that the salinity of this supply is itself increasing over time, and this does introduce the possibility that agricultural potential will be destroyed rather than enhanced. Once again aquifer recharge and increased rain have changed the situation but do not affect the basic structure of the model or its capability, in conjunction with local knowledge, to support the exploration of futures in response to these changes. The way in which the model captures these behaviours is, at each time step, to take each zone in turn and go through the following logical sequence (Figure 12–5). First, it asks whether the surface has sufficient water in it to allow growth of the crops. This presents a first logical gate: surface wet— no action, or surface too dry. In the latter case, the initial decision is to turn on the pumps and bring underground water to the surface, but this requires two further steps in order to decide what should happen. First, is there adequate water in the aquifer? If yes, then the pumps are turned on and the surface water is replenished. If not, then there is pressure to get canal water supplied. Another reason to ask for canal water concerns the surface water salinity which will reach a critical level when the water in the aquifer is saline. However, this water is not infinite, and indeed until 1994, there was only a limited supply available. This is shared out among those zones which it is decided to supply and may or may not be adequate for full production. There may be a resulting reduction in yield and this is allowed for in the model. Similarly, if there is no canal, and the aquifer runs dry, then there will be a shortfall of water with a corresponding reduction in the crop yield. In order to decide whether the farmers of a particular zone are pumping water, the model examines the following two series of questions at each time step. The first test concerns the need to turn the pumps on because the surface soil is too dry. The second case concerns the situation where


209

Figure 12–5 Water choice decision tree.

Figure 12–6 Canal water decision tree.

canal water is sought because the salt from the aquifer has already made the surface soil salty (Figure 12–6). This is a decision which faces the farmers when they have already allocated land to different crops. Over a longer time scale, they can modify their crop choice in the light of the reduced production of irrigated crops, but of course, if the salination of the surface is too strong then it will reduce the future potential of all possible crops. The model connects the region to the canal infrastructure and in doing so allows access to only a certain fraction of farmers. For geographic reasons, it may only be practical to deliver water to a limited region, and the others must therefore continue to make crop choices which rely on rainfall or pumped water from the aquifer. Another factor that the model takes into account is the limited capacity of a canal. As more areas demand access to the canal water, they may have to share a limited supply. The model assumes that the sharing is proportional to need, and that if extra water is required then this must be pumped from the aquifer. The increased economic production of the system must be weighed against the costs of building and maintaining different canal systems. The salinity levels of the Anavalos springs is also a matter of concern because while it appears that levels have been rising over time they have certainly been fluctuating. Different assumptions can be made in the running of the model, from the continuing increase in salinity to very high levels, to its stabilisation at present levels, or indeed to the introduction of desalination equipment at the

210


pumping station to supply fresh water. Once again, the apparent gains in production must be set against the costs involved in any such project. Indeed, an important aspect of the model must be to estimate the overall gains or losses taking into account the costs of subsidies and infrastructure. The strategic model of water flow presented in this chapter, in conjunction with the decision making model outlined in chapter ten, have described the development of a preliminary integrated dynamic model of agricultural land-use and water salt interactions for the Argolid Valley. The next chapter will consider some of the deficiencies in this approach and will discuss ongoing work that is being carried out to rectify them.

13. DEVELOPMENT OF AN ENHANCED INTEGRATED DYNAMIC MODEL Tim Oxley

INTRODUCTION The crop choice and water/salt flow components of the model described in the preceding chapters represents a preliminary development of an integrated dynamic model of the Argolid. Earlier chapters have described the complicated interactions that have defined the Argolid farming environment and the decisions relating to it. These have been greatly simplified in the preliminary model which omits many of the attributes that are necessary to provide a reasonable framework for policy exploration. Each of the key processes modelled so far (decision making, crop production and hydrology) would benefit from the inclusion of additional factors. The preliminary model of crop choice is based upon purely econometric representations which can adequately address the effects of financial influences such as subsidies on crops, the cost of water, and market prices although the impact of international market dynamics could be usefully added. The model also inhibits the rate of changeover between crops which are deemed desirable on an econometric basis but fail to reflect the other influences upon crop choice (e.g. duration of changeover). Finally, this econometric basis is unable to account for economically ‘irrational’ decisions based upon cultural influences and for example a possible transfer of emphasis between tourism and agriculture as the basis of farmers incomes. Overall crop production is estimated by the model each year on the basis of static productivity rates modified by the salinity of the soil, the soil type and the water availability. However, these estimates are made at the beginning of the year, for the purposes of decisions affecting the hectarage of each crop. They do not account for growing seasons and the potential of unexpected perturbations such as rainstorms affecting the harvests; determination of the actual production and its deviation from the estimates would considerably enhance the decision making for the following year. Finally, the hydrological dynamics revolve around the definition of the seven zones. This restricts the model to this single applicaiton and location (the Argolid) and does not adequately address surface water flows. This chapter addresses these failings in the preliminary model and outlines the approach currently being undertaken in the development of an enhanced integrated, multi-scale, self-organizing model for both the Argolid Valley and the Marina Baixa, region in Valencia, Spain. This will set the problems of resource degradation (water depletion and pollution, erosion an desertification) within a holistic framework of nested “response units”. The application to the Marina Baixa, where aquifer water does not present a problem, reflects a necessary step in the model development of a

212


slope model (see below) based upon data collected from complementary work into processes of soil erosion and the effect of vegetation upon hydrology.1 The generic nature of this enhanced model will subsequently allow for its transportation back to the Argolid Valley. MODEL DEVELOPMENT The amended model is being developed from the integration of existing models of hydrological flows across catchments (Allen, Oxley and Strathern, 1996) with the preliminary model of the Argolid described in the preceding chapters. Appropriate enhancements will have been made to the Argolid model in terms of aquifer flows and farmer decision making. The model will also include dynamics of crop growth, vegetation cover, soil porosity and composition, and erosion as described below. This reflects a deliberate attempt to integrate sub-surface and aquifer dynamics which are important in the Argolid, where there has been water depletion and salination, with the surface dynamics relating to supply and distribution that are more important in the Spanish case. Finally, the detailed approach to the representation of vegetation cover in the Marina Baixa will be incorporated into the generic model. Catchment Based Hydrological Model Previous work within the IERC had developed a dynamic and integrated catchment model of the Rhone Valley (Allen, Oxley and Strathern, 1996; Figure 13–1). This linked socio-economic factors such as the location and size of different populations and economic activities, to anthropogenic inputs to the river such as nitrate, phosphates and organic matter. The model describes the biochemical transformations that occur as the water carries these nutrients etc. down through the hierarchy of streams, and of sub-basins to the Mediterranean. It calculates the changing flows of water in the different branches of the river, as well as the concentrations of some seventeen different factors including oxygen, nitrates, phosphates, ammonia and organic matter, and is in good agreement with the observed data. This model allows for the exploration of the water quality in all the different branches and sub-basins, under a variety of conditions, including various scenarios of economic and demographic development. Similarly a range of different environmental policies can be explored (i.e. relating to Nitrate Sensitive Areas) as can changes in agriculture and other land uses, or indeed climate change. The model is structured to reflect both the hierarchical definition of the watershed, based upon catchments and sub-catchments, and the hierarchy of the socio-economic environment as defined by the administrative regions. This structure is represented schematically in Figure 13–2. The hydrological component of the model is driven by climatic data and geomorphological definitions of the individual sub-catchments, with nutrient/chemical inputs resulting from agriculture and human activity. The hydrological and socio-economic components communicate via an interface which relates the administrative regions to the sub-catchments, and vice-versa. Simulations are possible at the spatial scale of the watershed, the major catchments, or the many micro-catchments. 1 Studies of the Marina Baixa and the Argolid have been included in the second phase of the Archaeomedes project and the Marina Baixa is a central case study for the ERMES project on the ‘Environmental Response of Mediterannean Systems’ (Ermes, 1997).

DEVELOPMENT OF AN ENHANCED INTEGRATED DYNAMIC MODEL

213

Figure 13–1 Hierarchical structure of the catchment based hydrological model developed for the Rhone Valley.

Aquifer and Farmer Decision Making Model The Argolid model described in previous chapters consists of two integrated sub-models. The first is a physical model of the water movement (Chapter twelve) and the second is a farmer decision making model resulting in the selection of particular crops based upon perceived profits and welfare (Chapter ten). A schematic representation of these models is presented in Figure 13–2. As was described above the aquifer model details the movement of water through the Argolid taking daily timesteps and each spatial zone in turn. It considers the net flows of water between the surface layer, the porous layer below it, and the underlying aquifer. In this way, the patterns of natural rainfall and evapotranspiration, soil permeability and slopes are linked with the movements of water through the ground and in the aquifer. This in turn relates to the changing patterns of hydraulic pressure which result from the varying actions of farmers in the different zones pumping irrigation water. This behaviour has meant that the aquifer in the Argolid, prior to 1995, had been severely reduced in size, become increasingly saline, and the productivity of much of the region had been restricted. The Farmer Decision Making model has drawn on interviews with farmers to determine the costs and benefits they consider, the risks that they run, their responses to those risks and the rate at which they can respond to changing opportunities. Information was also gathered concerning the costs involved in farming the different crops, and in moving from non-irrigated to irrigated crops.

214


Figure 13–2 Schematic representation of the structure of the Aquifer and Farmer decision making model developed for the Argolid.

The model may thus calculate profits that could be expected for a specific crop in a particular zone, given the labour involved, the water costs or costs of drilling, seed and fertilizer costs and those for electricity and transport. This calculation also accounts for the variation in yields that may be expected under different soil types and growing levels of salination. The integration of these two models enabled the exploration and analysis of various policy options involving factors such as market pricing, subsidies and irrigation costs. However, as discussed above, the model also displays certain shortcomings which should be rectified by the developments described in this chapter. THE ENHANCED GENERIC MODEL The enhanced, generic model described in this chapter is being simultaneously developed from the integration of these catchment and aquifer/decision making models for both the Argolid Valley and the Marina Baixa region of Spain. It will also incorporate significant additional factors resulting from soil nutrient composition, crop growth and erosion, and will provide an interactive environment within which the exploration of policy options by stakeholders will be facilitated. The overall structure of the proposed model is presented in Figure 13–3. This shows the main linkages between the three sub-models discussed above, and the development of a Crop/Soil submodel using aspects of the Farmer Decision Making model and related work concerned with the retention and/or release of nutrients within the soil system (Oxley, 1994). it will also utilize data relating to the soil composition, crop growth and erosion. The Catchment sub-model will be driven by definitions of the geomorphology and topography of the region, and by meteorological and demographic data. Further development of this sub-model will include losses of water to the aquifer from the rivers, similar losses through the crop/soil profile, and the abstraction of water for both agricultural and urban uses (including tourism). The calculation of nutrient inputs to this sub-model will be controlled by the Crop/Soil sub-model. The Catchment sub-model will also incorporate information about water treatment plants, their location within the sub-catchments and their relationship to population concentrations and water quality. The Aquifer sub-model will be further developed to represent the regions using a regular spatial grid, as opposed to the seven unique zones defined in the preliminary model developed for the Argolid. This sub-model will then be driven, not directly by rainfall and evapotranspiration, but


215

Figure 13–3 Structure of the enhanced, integrated catchment, aquifer and decision making model.

indirectly through the Catchment sub-model (surface water) and the Crop/Soil sub-model (subsurface water). Pumping from the aquifer will result in the appropriate transfer to either surface or soil water, or loss through human consumption. A spatial interface will be developed which relates the spatial grid, upon which the aquifer is defined, to the natural, irregular definition of subcatchments across the regions. The Crop/Soil sub-model will be developed from that part of the Argolid Farmer Decision Making model which relates to crop growth and production, from the dynamics of nutrient retention and release by the soil and from the dynamics of soil erosion in the regions (ERMES, 1997). This submodel will act as a buffer for both water and nutrient movement between the surface and the aquifer. The model will represent spatial variation of phenomena using a regular spatial grid within individual sub-catchments. This representation will be linked to both the spatial definition of the catchments and the grid defining the aquifer. The spatial representation of cropping patterns will be driven by the Farmer Decision Making submodel, and will be used for the calculation of soil erosion, surface-water flow rates, water demand and usage, and nutrient uptake from the soil. This decision making sub-model will, together with the existing factors such as pricing and subsidies on crops, incorporate the hierarchical structure of the farming communities (made up of individual farmers, the landowners and farming cooperatives) defining the spatial levels at which policies may affect the agricultural communities and at which individual and collective responses may be evident. Thus, the Farmer Decision Making submodel will drive the Crop/Soil sub-model through decisions on crops and hectarage, and influence the hydrological sub-models through the subsequent determination of water demand. The anticipated data requirements for the integrated modelling activity and the sources of that data are shown in Table 13–1.

216


Table 13–1 Data requirements for the development of integrated model.

These data, and more, will provide the basis of an integrated database which will be accessed by each sub-model as required. The collection of the data has also started to establish an historical perspective about the use of land and natural resources in both the Marina Baixa and the Argolid regions. The computational structure of the Marina Baixa model will reflect the schematic representation shown in Figure 13–4 where each sub-model is treated as an individual object communicating with other objects through and under the control of a meta-level driver. This allows for the parallel and semi-independent development of each sub-model using a high level driver with a user interface for overall model control, the definition of policy scenarios, and direct interaction with selected areas of the integrated database.


217

Figure 13–4 Computational structure of the generic, integrated catchment, aquifer, crop/soil and farmer decision making model.

A number of points can be raised at this stage. Firstly, it is anticipated that the model itself can generate certain types of information based upon secondary data sources. For example nutrient levels and their transportation can be estimated through a knowledge of the bio-physical features of the landscape (soil type, vegetation, slope, rainfall etc.) in conjunction with information about the application of agricultural inputs. This highlights the practical value of social enquiry techniques for providing data about bio-physical systems quickly and across a wide area but at lower levels of resolution. This approach inevitably benefits from the data acquired through existing monitoring sites for validation. Secondly, the need to include stakeholders in the description, specification and interpretation of the models as part of a policy relevant method has already been discussed. This is represented in Figure 13–5 as the feedback between the model output (interpretation) and its inputs (description and specification) which provides the cultural context for exploring different policy options. What is less clear is the extent to which cultural variability can be represented. For the time being it is assumed that the ‘lens’ through which different futures are explored by stakeholders will result in insights that take account of those differences while not expecting them to be reproduced unless some appropriate proxy measure is available. For example status and quality of life are both important aspects of rural life, however, not only are they difficult to break down and value they also mean different things to different people at different times. These differences can be more reliably observed through the interaction with, and the exploration of, different policy options generated through the model. Figure 13.5 provides a schematic representation of the integrated model which

218


Figure 13–5 A schematic representation of the spatial mapping of data required for integrated, generic model of the Argolid or the Marina Baixa.

highlights the role of the user and thereby the ability to use the model with culturally, and geographically, diffuse stakeholders.


219

A number of ‘interactive’ and iterative stages will be carried out in the course of model development. Firstly the existing model will be explored with individuals who are ‘local’ to the case study areas. This is intended to expose any obvious weaknesses which can be rectified alongside the development outlined above. A further iteration will then involve the model being exposed to representatives of stakeholder groups. This will examine both the quality of the interface and more importantly the diagnostic relevance and utility of the model as an exploratory, not a predictive, tool. Once again weaknesses in the model as it represents local processes will be identified. A third iteration will involve the piloting of the model with a range of stakeholders over a more prolonged period of time and consideration of the reflexive nature of the model-actor-interaction. The preceding discussion describes the theoretical approach to modelling the Marina Baixa and the Argolid and the underlying structure of the generic integrated model. The development of the individual sub-models is ongoing and will reflect an enhancement of the aquifer and decision making model described in the preceding chapters combined with an adapted version of the catchment and a dynamic model of slope, soil, vegetation and water which is described in more detail below. This novel model of slope and erosion dynamics will ultimately link precipitation, rivers and the aquifers; the lack of which a shortcoming of both the catchment model and the original Argolid model. Data has been collected describing the precipitation patterns, microclimates, hydrology, water consumption and water treatment of both the Marina Baixa and the Argolid, along with information about demographic changes and land-use (Table 13–1). These data, together with the geomorphological descriptions of the individual sub-catchments, provide the basis for calibration and verification of the catchment sub-model. The individual sub-catchments have been defined spatially, with the detail of stream lengths, slope and related drainage areas still being determined. This spatial definition of the sub-catchments is being examined against the spatial definition of the administrative regions in each study area in order to derive an interface between these two representations. This will facilitate the use of data defined by administrative regions to be incorporated into the model which is driven by the catchment definitions. These spatial definitions represent the topographical and demographic layers, respectively, in Figure 13–5. The spatial definition of the aquifer model will provide another of the layers shown in the figure (with a spatial grid appropriate to the scale of the entire region), and finally, the crop/soil model will be mapped onto this spatial structure, but with a spatial grid which reflects the scale of individual sub-catchments rather than the entire watershed. Having defined this spatial hierarchy, within which each of the sub-models will operate, the watercourses, reservoirs, canals and other hydrological features (wells and pipelines etc.) can be situated within the model environment. Figure 13–6 indicates the major watercourses and water distribution network in the Marina Baixa and Figure 13–7 shows the general topography and river network in the Argolid. It should be noted that many of the individual rivers and streams in the Argolid rarely contain water, although they define the course of the rivers during rainfall events. This completes the description of the enhanced, integrated and generic model and the discussion relating to the interactions of individual sub-models within the spatio-temporal context of these two watersheds (Marina Baixa and the Argolid), except for the dynamic model of slope, vegetation and water, which is described in detail below.

220


Figure 13–6 The main watercourses and water distribution network in the Marina Baixa.

Figure 13–7 The major watercourses (excl. canals) and topography of the Argolid, Greece.

A DYNAMIC MODEL OF SLOPE, VEGETATION AND WATER The following section presents a simple model which is incorporated within the crop soil model (see Figure 13–4). This is designed to show how the mechanisms that operate between vegetation cover, water storage capacity, water infiltration, evaporation and run-off can determine the spatial patterns


221

Figure 13–8 The positive feedback loops that impact upon vegetation.

of vegetation. These demonstrate a complex, non-linear response to different average amounts of rainfall, slope and aspect, and provide a base for the dynamic linkage of multiscalar phenomena. Some detailed models have been developed of these phenomena and a paper by Imeson et al. (1996) presents a spatial model which links patch dynamics to the more macroscopic Desertification Response Units. However, the model is only one dimensional, and the non-linear mechanisms proposed are not necessarily the appropriate ones. Here we shall take a scheme of interaction that is generally accepted as being correct, and show that this can lead to the self-organization of Response Units of macrostructure at a spatial level above that of the small patches in interaction. The interaction diagram for the model developed for Marina Baixa and the Argolid is shown schematically in Figure 13–8. The following calculations for the model are structured hierarchically. For example biomass and storage capacity are assumed to change only on a daily basis, while rainfall and patch conditions concerning wetness and water infiltration can change over hours. The interactions shown in Figure 13–8 exist in each spatial zone which are connected by flows of surface water. The first calculations are made with a daily timestep:

Biomass: where xi=biomass of vegetation at i, Storagei=water stored in the soil at i,

222


storxmax=Maximum density possible in zone, b=growth rate of x, m=decay rate of x, Creep=rate of sideways diffusion. In each zone, there is a surface layer, with a storage capacity located below related to the amount of organic matter that has grown and decayed in the zone. So, it is the decay rate, mx, that produces the organic matter that acts as a water storage capacity in the subsoil. It is assumed that this capacity, maxstoragei (where i is really a point on the slope surface and has two spatial indices), decays naturally at a certain rate, but is increased according to the equation: which describes a 5% annual decay rate, and an increase of 10% of the biomass decaying in the zone. In considering the water flows through the system, we move to a finer timescale of hours in order to capture the effects of irregular rainfalls, and the non-linear effects of these on the system. Considering the density of water on the surface, then:

where rain, is the instantaneous rainfall at i (per hour), evt is the evaporation rate, and infini is the rate at which water can infiltrate at i. So the change in surface water per unit area depends on the rain falling, on the water evaporating, or infiltrating into the subsoil, as well as the net flows of water in and out both along and across the slope. The change in water storage at i, k is given by:

where infouti is the rate at which stored water can leak out to recharge the groundwater. From these equations, we can calculate the average run-off per unit area, the average evaporation per unit area, and the average aquifer recharge per unit area for a given average rainfall, slope, and aspect. We can choose how many zones to model, and for the initial simulations a 10×10 grid has been chosen (Figure 13–9). The model shows how the positive feedback loop expressed by the development of storage capacity and of vegetation cover leads to a clear pattern of either presence or absense of vegetation in the different zones. Either the vegetation succeeds in developing sufficient storage capacity to survive the dry period, or it doesn’t. Figure 13–9(c) and (d) show the effect of reduced rainfall. In Figure 13–10(a) a simulation of a 40×8 slope is shown. This starts from an initial condition of a random scattering of biomass, and allows the positive feedback loop of vegetation and storage capacity to structure the system. It does so by creating very simple “response units” which consist of zones with biomass, and zones without. In other words, the condition that plants survive on this slope is that they be able to “capture” a territory greater than a single zone. In the model there are weak flows allowed horizontally as well as the strong flows on the slope. Clearly, then, the model shows that a “functional structure” emerges in which patches of vegetation survive. This idea can be verified by repeating exactly the experiment above, but instead of starting from an initial condition of a random scattering of biomass, with average density 1, we shall start with a uniform initial condition of x=1 in each zone.


223

Figure 13–9 Response units of vegetation patches develop after 10 years on this 10×10 spatial grid, with rainfall set to (a) 200mm, (b) 200mm, (c) 150mm and (d) 100mm.

Figure 13–10 A 40×8 slope after 10 years (a) random biomass spread develops response units of vegetation, (b) Uniform biomass fails to develop response units, and (c) Some response units emerge as a result of spatially uneven rainfall.

As we see from Figure 13–10(b), after 10 years the same system fails to develop any biomass, and runoff is much larger. This is because the uniform distribution of biomass does not allow any particular zones to develop sufficiently, and to “supress” the development of neighbouring zones, thus providing themselves with a wider spatial area to draw water from. In this model, the rainfall is itself usually allowed to be randomly distributed both in time and space. The temporal randomness corresponds to the idea that in winter there are frequent shower events, and in summer very few. We have retained this in Figure 13–10(c), but we have suppressed

224


the addition spatial randomness of the rainfall. If we allow this spatial randomness to occur, but start the biomass from a uniform distribution, then although it is not as effective as in (a), some response units do emerge simply as a result of the symmetry breaking due to the rainfall distribution. Some further exploratory simulations were carried out, this time based upon a 3-dimensional slope in preference to the slopes described above. These show a landscape of two hills with a saddle point between them, with the left hand hill slightly higher than the right (see Figure 13–11 (a)). This surface is initially made up of ‘removable’ soil overlaying a rock core which is not erodable in the timescale of the simulations. The vegetation is initially randomly scattered and the initial flows are small.

Figure 13–11 (a) Initial contour map, (b) Contour map after 40 years, (c) Vegetation map after 40 years, and (d) the flows concentrated in the gullies between the vegetation and bare rock.

The simulation proceeds and the model allows vegetation to flourish if there is water in the soil, and the storage capacity enables survival through the dry season. The storage capacity is increased by the biological activity of the vegetation, and water flows depend upon the relative height of neighbouring cells. The flows carry some of the ‘removable’ soil to neighbouring cells where it is deposited, simulating erosion and subsequent deposition. Figure 13–11(b) shows the soil becoming deposited in the saddle point and spilling between the two hills.


225

Figure 13–12 The relatively stable situation encountered after running the model for 200 years.

After some forty years, the profile of the hill has changed as the result of soil eroding from the steep outer sides and being lost, hills. Gullies form around the raised lower slopes, where vegetation has stabilized the soils. The vegetation has managed to survive on the eroded slopes. Patches of vegetation still survive on the otherwise bare outerslopes of the peaks, but in general these are bare. The eroded slopes, however, between the peaks have been stabilised and the water flows are now channelled around the lower slopes. The vegetation on the deposited soil clearly leads to low overland flows on these surfaces, although, the gullies concentrate flows. However, the bare, steep and rocky outer slopes have also high flows. Running the model for a long period, in this case 200 years, then the relatively stable situation emerges shown in Figure 13–12 emerges. Finally, ongoing work remains into the development and integration of the individual sub-models. Additionally, the collection of further hydrological and biophysical data is necessary to complement that which has been collected from social enquiry into farmer responses. The preliminary results shown above indicate the importance of representing the complicated interactions between vegetation, soils and hydrology when considering the landscape within a typical catchment. The slope model will be developed further to establish the run-off to the smallest stream

226


orders, and the evaporation and aquifer recharge that will be expected according the soil type, vegetation cover, land use, slope and aspect. In this way, consistent up-scaling will be possible so that the effects of policies and changing land use on the microprocesses of the soil and of plants can be traced through to the catchment as a whole. Thus, the integrated model described in this chapter will include the effects of erosion in these hierarchical models. So, the surface flows will transport sediment, according to some non-linear relationship with velocity, and this sediment will be deposited further down the slope, increasing the soil depth there and the sediment available for transport. The non-linearities will produce gullies and ridges in the flow patterns, and the vegetation will reflect soil depth and storage capacity. In this way we will be able to build up dynamic models which, on the small scale, change and shape the geomorphology, and lead in the longer term to changes in the hierarchy of sub-basins making up the major catchments. From this, long term estimates of surface water availability, groundwater recharge, and of erosion and flood risks will be generated, allowing policy exploration for the effects of climate and land-use changes. These will be considered further in the final chapter.

14. WHERE TO FROM HERE? Mark Lemon and Tim Oxley

The final few pages of this text will summarise what is meant by policy relevant and integrative research and will consider a number of the current activities being undertaken by the team to clarify and develop the method. This book has attempted to address some of the tensions that inevitably occur between science, policy and the world as it is interpreted and negotiated by different actors-including scientists and policy makers. Knowledge generated by scientific research can be difficult to assimilate and exploit, it also tends to be determined by the agenda and interests of the scientist and the body funding the work. However if research is constrained to addressing ‘useful’ topics then the creativity that often leads to new insights can also be restricted. Further confusion can arise over the distinction between research which is designed to evaluate policy, and to characterise appropriate policies (policy research), and that which aims to explore, diagnose and analyse complex situations in order to provide information which identifies the options for policy formulation (policy relevant research). By focusing upon the range of responses and interpretations at the local level policy relevant research moves from the disaggregate to the aggregate i.e. from the bottom up, and includes perceptions of the policy formulation and delivery process in that analysis. Policy and policy research, in common with much science, invariably adopts a top-down approach which is grounded in the aggregate and applied to the disaggregate. This text has sought to examine the impacts of the policy process at the local level and has advocated the need for it to be more aware of the diversity at that level as an indicator of its impacts and a determinant of more appropriate policy formulation and delivery. The policy-relevant approach therefore invariably entails the development of transdisciplinary mechanisms to characterise situations and phenomena that are seen to constitute a problem at the local level. This approach is driven by issues as they are perceived in the ‘real’ world. It emphasises a number of skills (e.g. social enquiry, modelling and soft complex systems thinking) which underpin a ‘transdisciplinary’, and integrative, paradigm through the facilitation of an effective cross disciplinary dialogue. Integrative research, therefore, requires that representatives of single disciplines retain their distinctive, and consistent, frames of reference while establishing an effective framework for communicating with each other. This is particularly relevant with disciplines that focus upon natural and social phenomena as discrete and independent without providing handles that allow us to interpret the interaction between them. Underpinning the development of these transdisciplinary skills, and the acceptance of systems as complex and subject to multiple interpretations, is the need to move away from the desire to predict and towards the capacity to adapt. In terms of methodology this means the exploration of a range of possible futures rather than the anticipation of any specific future path. This has particularly

228


important implications for how we interpret sustainability in that it encourages us to look towards viable paths rather than end states (Park, 1993). In this context sustainability is a process and the identification of multiple pathways towards what appears to be unsustainable is a more appropriate approach than focusing upon the achievement of specific ‘measures’ of sustainability. A second point arising from this unwillingness to predict is that it runs counter to the role which is often prescribed for science by policy and politics. This is particularly important with the production of ‘deliverables’ such as computer based models and the translation of scenarios into futures; they can quickly become justificatory rather than exploratory. This argument is not intended to give the impression that just because the future is uncertain and does not lend itself to prediction we are excused from anticipating it. Indeed, as has been repeated throughout this text, the exploration of possible futures, based upon an appreciation of diversity at the local level, is fundamental to adaptive policy processes. At the same time, decisions about uncertain futures should not be based upon the impression of scientific certainty; responsibility lies within the political arena. Therefore while science and politics are part of the same complex processes their responsibilities must remain clear and distinct. In summary, the central message of this book has been the inseparability of natural and human systems and an acceptance of the need to adapt rather than predict. Underlying this, and the policy relevant component of the work, is the argument that environmental concerns are grounded in issues and not disciplines. Therefore, in order to clarify what issues are salient it is essential to identify, interact with and understand how stakeholders perceive such processes. These multiple interpretations contribute to the inherent noise of ‘complex systems’ which cannot be removed through the adoption of a reductionist lens. This requirement to identify multiple perspectives, decisions and behaviours, and to incorporate them into our representation of a system, in turn highlights two further issues for consideration. Firstly, it is questionable how far we can incorporate the ‘cultural’ and qualitative aspects of human interactions and behaviours into a mathematical representation. In the research reported above decisions have been represented through the response to a limited number of variables (e.g. changing water price) and the behavioural manifestations of those decisions (e.g. amount and variety of crops grown). Our understanding of these decisions, and the behaviours relating to them, is somewhat more sophisticated and is based upon the identification of qualitative phenomena that combine to restrict or expand options (e.g. by considering the social status attached to farming and the likelihood of growing crops that are labour intensive). The jury is out as to how far we can go towards a computerised representation of cognitive processes. However, by establishing a clearer picture of what factors influence decisions and behaviours we are also creating a framework for interpreting the output of more simplified computer representations. This is a fundamental point in policy relevant research because it argues the need to establish what phenomena influence behaviour and the role to be played by such insights for evaluating the range and distribution of responses to different policy interventions (e.g. removing subsidy or price support). The same point can be extended to the policy process itself in that effective policy formulation must take into account what issues are relevant, and how different policies will be interpreted and responded to, at the local level. The delivery of policy has also been seen as fundamental to the options that are perceived locally. The interpretation of policy instruments by the administration, and the effectiveness of that administration in managing their delivery, influence the efficiency with which information is provided to stakeholders (e.g. farmers) and the message that is communicated. Therefore it is not sufficient merely to focus upon local interpretation and responses as indicators of policy impact.

WHERE TO FROM HERE?

229

Analysis of the policy implementation process itself, and the agencies involved, can provide invaluable insights into the (mis)match between the objectives of policies as they are formulated and their subsequent impact. FUTURE DEVELOPMENTS TO INTEGRATIVE METHOD In the preceding chapters we have described firstly a strategic model of water flow, incorporating some simple econometric dynamics, and secondly the development of an enhanced, integrated dynamic socio-economic, biophysical and hydrological model which overcomes many of the shortcomings of the former. The continued development of the integrated model will address two specific areas: the incorporation of erosion dynamics into a slope model to complement and enhance the existing dynamics of water flows and vegetation growth, and the derivation of ‘response units’ relating to farmers’ decisions concerning crops and determined from a series of characterisations of both the farmers and their cultivatable land.

Figure 14–1 Additional processes to be incorporated into the slope sub-model.

The additional processes which are incorporated into the slope sub-model to represent the dynamics of erosion as a result of overland water flows, slope and vegetation cover are shown in Figure 14–1. The effect of these processes is to enable the soil to be removed from a given cell in the model by the water run-off (Erosion) which will subsequently be deposited in an adjacent cell in relation to the water run-in. This is equivalent to the total run-off from the adjacent cells—or into

230


streams and deposited as sediment if the run-off results in channel flow. The soil depth, below which is deemed to be porous or non-porous rock, is recorded for each individual cell in the model and will reflect the rates of erosion and deposition for the given cell. The dynamics which emerge from this model of water, vegetation and erosion on a slope will be initially calibrated using existing data relating to the effects of rainfall events and vegetation cover on the dynamics of erosion. These data have been collected from three complementary field sites in the Alicante region of Spain (Calvo-Cases et al., 1995). More comprehensive calibration may be possible using archaeological records of erosion and river flows in the Rhone valley. This calibration will provide validation of the integration of the slope model with the catchment river model which was initially adapted from the Rhone. Such calibration in the Marina Baixa and the Argolid would be problematic since there is negligible water in the rivers of these two areas except following shortterm rainfall events; the disparity between the geomorphological time scales associated with erosion and the ‘instantaneous’ rainfall events is too extreme. The second area for the continued development of the integrated model is the derivation of ‘response units’ relating to farmers’ decisions concerning crops. These are determined from a series of characterisations of both the farmers and their cultivatable land. The enhanced decision making sub-model will be based upon both the original model of the Argolid (Allen et al., 1995) and a nested master-equation model of crop choice dynamics based upon a hierarchical definition of the various criteria involved in farmers’ decision making (Winder et al., 1997). In order to derive the appropriate characterisations of farmers and their land the region would be divided into a grid based upon the resolution of the slope/erosion sub-model. Each individual cell in the grid may then be characterised, through the utilization of spatial overlays, so that every cell may be defined by the vegetation cover—using local survey data or 100m land cover data assimilated by the EEA (1997)—slope, aspect, soil type, geology, the existence of terracing and irrigation (including the type of irrigation and the quality and source of the water), crop type (highlighting the annual or perennial nature of crops), and the type of farmer (i.e. full or part-time, organic or non-organic). Based upon characterisations which highlight these specific decision making criteria (so far as crop decisions are concerned), ‘response units’ which may be. expected to reflect similar decision making dynamics will begin to emerge. Two such spatial characterisations, relating to the aspect and slopes of a 6 km square region of the Marina Baixa at a resolution of 100m, are presented in Figures 14–2 and 14–3. With such information it is possible to build up an understanding of how a farmer may reach decisions concerning crop type, the requirement for irrigation and the need for terracing. Different crops may be suited to northerly or southerly aspects, whereas terracing would involve unnecessary effort and expense in a region where the slope is negligible. Furthermore, as has been discussed in preceding chapters, farmers in areas where there are perennial crops such as Olives, Oranges and Medlars (the dominant crop in this area of Spain) would display decision making characteristics on different time scales to those cultivating annual crops. Decisions would relate to water and other crop specific requirements as opposed to decisions to plant tomatoes one year and another vegetable the next. Two Medlar farmers may well respond differently if one was full-time and the other part-time (with additional income generated from tourism based activities). Since Medlars are labour intensive, the part-time farmer may choose to grow Oranges instead, a situation reminiscent of the situation encountered in the Argolid. It is clear, therefore, that by spatially characterising the area being modelled in this manner characteristic ‘response units’ will rapidly emerge reflecting many of the farmers’ criteria for decision making, both throughout the growing season (i.e. relating to water requirements) and annually

WHERE TO FROM HERE?

231

Figure 14–2 Characterisation of a 6 km square area of Marina Baixa by aspect at 100 m resolution.

(relating to changes in crops). The completed integrated model will thus be able to address the spatially and temporally disparate processes surrounding the inter-relationships between the farmer and the natural environment, facilitating the analysis of various policy options and their effect upon crop yields, erosion dynamics, hydrological flows, water quality and other factors of significance to the farmer. Retracing Policy Impacts The models outlined in this text have provided a provisional platform for the exploration of the effect of policy on the co-evolution of the water and crop systems and the agricultural economy. The previous section outlined the development of these models in terms of their potential for integration. This integrated ‘suite’ of models can be used to support a number of ‘experiments’ which ‘re-run’ the historical representation of the system with changes in key variables. The objective of this exercise is to explore the way in which policy might have avoided the emergent economic and environmental problems described in the Argolid. The initial phase of the exploration is relatively simple in that the ability to manipulate key variables is incorporated into the structure of the model (i.e. changes in crop price, water costs) and the subsequent running of the model will represent the co-evolution of the agricultural economy with water availability and salinity. This approach

232


Figure 14–3 Characterisation of a 6km square area of Marina Baixa by slope at 100m resolution.

deliberately reverses the normal perspective on modelling. We know what has happened but not how particular paths might have been avoided or the negative impacts experienced in the area, restricted. The generation of ‘new histories’ will facilitate an exploration of the relative effects that different types and scale of policy intervention might have had. It must be remembered that these are explorations of alternative trajectories and not representations of what should have happened. A key element of policy relevant research has been seen as the ability to observe and interact with key actors, groups and agencies in order to elicit what issues are salient to them and how they interpret, and respond to, issues that are determined elsewhere e.g. by other stakeholders or in response to policy or scientific agendas. In the Argolid case study much of the field work to obtain this information was undertaken by local researchers who knew the geography and culture of the area. They also had an agronomic background which meant that they could discuss specific farming issues with the respondents. The local knowledge of these researchers was subsequently drawn upon to determine the spatial zones for modelling representation and to suggest the range of responses to some of the computer generated scenarios. Ongoing work is taking this process one stage further by involving local representatives and policy administrators in the interpretation of modelled responses to different policy interventions. As has been seen this interpretation could refer to the range of possible futures; it could equally apply to the development of different historical trajectories that

WHERE TO FROM HERE?

233

might have been generated by policy intervention (i.e. what is the distance between the present and other contemporary representations that have emerged as a result of different policies). MODELS, METHODS AND USERS—A DIALOGUE? A number of key methodological questions have been raised in this pursuit of a transdisciplinary approach for exploring complex environmental issues. One aspect of that complexity is that it is subject to multiple interpretations, and what is considered important varies not only between individuals and groups, but their evaluation of an issue may well change through time. The scope of what is possible will also vary according to the available information and the cultural context within which that information is interpreted. This has been seen as central to the difference between decision space (what is perceived to be possible) and opportunity space (what is possible). An implicit criticism of much of the policy process, and indeed of science, has been that it has paid too little attention to the mis-match between the perception of issues and their ‘objective’ representation based upon an assumption of perfect information. However, in the same way that high resolution data cannot usually be obtained to provide an overview of the physical environment (e.g. of an area’s hydrology), perceptual and ethnographic minutiae is seldom available to support such a picture of the cultural and socio-economic conditions. In both cases specialist information is fundamental to understanding the detail but the dynamic nature of environmental issues, and the policies relating to them, can often mean that such insights arrive too late to support or inform adaptive responses. Two methodological observations arise from this argument. Firstly we need a medium which can provide a simplified representation or model of a complicated landscape. Secondly, representation must incorporate the different perspectives of those who are making decisions which impact upon that landscape. Figure 14–4 suggests that the specification of a model should incorporate different local interpretations (e.g. lenses A and B) of an environment. As was discussed in chapter nine these interpretations are different configurations of bio-physical and socio-economic and cultural factors. To an extent, this specification process has already been included in the method described above through the semi-structured and structured interviews, and the use of local researchers to interpret the findings. What has not been established to date is a mechanism whereby local actors, apart from the researchers, are involved in the interpretation of the modelled output. By including local actors in the process of articulating and interpreting the model it is possible to conceive of an iterative process whereby the model is amended in response to new input (modelmodel1 in Figure 14–5). Similarly where the same actors (A and A1) are interpreting the modelled output it is conceivable that their view of the world will be altered, albeit in a minimal way. It is this reflection that can form the basis for policy relevant exploration where different futures, and pasts, are explored and perceived responses considered. The generation of futures can also be interpreted using more in depth knowledge from the physical and ethnographic sciences. While it may have been difficult to incorporate such insights into the specification of the model itself the scientific lens can be usefully employed to explore and interpret the output. This view of integrative method immediately questions the role of science as an objective observer and analyst and recognises the reflexive nature of the process whereby the scientist is inseparable from the focus of the study. In the context of the Argolid case study this was important not only because much of the technological intervention was based on scientific expertise and guidance, but

234


Figure 14–4 Models, methods and users—a dialogue 1.

Figure 14–5 Models, methods and users—a dialogue 2.

WHERE TO FROM HERE?

235

because when the related decisions and behaviour did not conform to the requirements of this technological model, there was limited understanding about what lay behind such ‘irrational’ behaviour. A FINAL THOUGHT: HOIST WITH OUR OWN PETARD? The completion of this text has been delayed because many of the phenomena that were identified as important in the first phase of the Argolid case study have undergone considerable change. In consequence a great deal of re-writing and updating was necessary, not only in terms of reporting the substantive changes themselves (e.g. more rainfall, reduction in orange subsidies), but because the story only made sense if these changes were put into a wider, systemic context. This is very much a case of being hoist with one’s own petard whereby the research itself was subject to the uncertainties of what was being investigated, and in order to respond to these changes, had to be flexible and adaptive. It is interesting to note however that some of the initial observations have supported this re-evaluation. For example the discontent that was expressed by some of the peripheral farmers towards those who were farming in the valley was based on the latter’s monocropping of price supported crops and the related heavy use of water which had led to pollution in the valley and depletion elsewhere. While the rainfall of the past two to three years has addressed, at least for the time being, the depletion problem and the reduction of price support has restricted access to what was perceived as easy money many farmers in the valley are not flexible enough to diversify their cropping (e.g. because of time constraints due to other work, a reluctance to undertake physical labour, lack of experience with alternatives). At the same time those full time farmers who have worked much of the periphery, in a more diverse way, have continued to farm in this way, but with a reduced concern over water. The lesson to be learnt is, however, a salutary one in that while it has been argued that policy must recognise that while complex processes do not fit neatly into administrative structures or predetermined geographical and temporal scales, exactly the same warning has to accompany research designed to improve our understanding of these processes. In the same way that policy has to look beyond its stated objectives and towards the processes by which it is delivered, and the multiple interpretations which exist at the point of delivery, so must policy relevant research look to itself and the manner in which it impacts upon, and adapts to, changes in the phenomena that are being investigated.

236

REFERENCES

Ackerman, F., Cropper, S. and Eden, C. (1991) Cognitive mapping for community operational research research —a user’s guide, Operational Research Tutorial Papers, pp. 37–51. Ackoff, R. (1970) ‘A Concept of Corporate Planning’, John Wiley & Son, London. Ackoff, R. (1983) ‘Beyond Prediction and Preparation’ Journal of Management Studies. 20(1), pp. 59–69. Adams-Weber, J.R. (1979) ‘Personal Construct Theory: Concepts and Applications’, John Wiley, Chichester. Alchian, A. (1950) ‘Uncertainty, Evolution and Economic Theory’, Journal of Political Economy, 58, pp. 211– 221. Allaire, Y. and Firsirotu, M. (1989) ‘Coping with Strategic Uncertainty’, Sloan Management Review, 30(3), pp. 7–15. Allen, P. (1992) Modelling Evolutionary and Complex Systems, World Futures, 34, pp. 105–123. Allen, P. (1996) Cities and Regions as evolutionary complex systems, Geographical Systems, 4, pp. 103–130. Allen, P. (1997) Cities and Regions as Self-Organizing Systems: Models of Complexity, Gordon and Breach, Reading. Allen, P. (1997) Evolving complexity in social science, in Altman, G. and Koch, W. (eds) New Paradigms for the Human Sciences, Cambridge. Allen, P. and McGlade, J. (1987) ‘Modelling complex human systems: A fisheries example’, European Journal of Operational Research, 30, pp. 147–167. Allen, P., Lemon, M. and Seaton, R. (1995) “An Integrated Physical and Socio-Economic Computer Model of the Argolid.” IERC, Cranfield. A contribution to the Archaeomedes Project EU-DGXXI (EV5V-0021). Allen, P., Lemon, M. and Seaton, R. (1996) “An Integrated Physical and Socio-Economic Computer Model of the Argolid. The Argolid Case Study.” A sub-project of Archaeomedes, DGXII Environment Programme. Allen, P., Lemon, M. et al. (1995) “Agricultural Production and Water Quality in the Argolid Valley, Greece: A Policy Relevant Study in Integrated Method”, Archaeomedes Report 1995, DGXII Environment Programme, Project EV5V-0021. Allen, P., Oxley, T. and Strathern, M. (1996) “A dynamic spatial model of water quality in the Rhone Valley” in [‘Coupled Ecological and Socio-Economic Modelling in the Rhone basin (CEcoSEcoM)’ CEE-DGXII Environment Programme, Project EV5V-CT94–0547]. Anderson, J.R. (1974) ‘Sparse Data, Climatic Variability and Yield Uncertainty in Response’ Analysis, American Journal of Agricultural Economics, 55, pp. 77–82. Antle, J.M. (1983) ‘Sequential Decision Making in Production Models’, American Journal of Agricultural Economics, 65(2), pp. 282–290. Arber, S. (1993) ‘The research process’, in Gilbert, N. ‘Researching Social Life’ Sage, London. Argolid Association of Agronomists (1992) ‘The Water Problem of the Argolid’ (in Greek). Ashby, W. (1958) General Systems Theory as a new Discipline, General Systems, 111, pp. 1–6. Bauer, S. and Kasnacoglu, H. (1990) ‘Nonlinear Programming Models for Sector Policy Analysis’, Economic Modelling, July, pp. 275–290. Beer, S. (1966) ‘Decision & Control.’ John Wiley & Son, London. Bell, C. and Newby, H. (eds) (1977) Doing Sociological Research, Allen and Unwin, London. Bellman, R.E. and Hartley, M.J. (1985) The Tree Crop Problem, Report No. DRD134, World Bank, Washington. Belous, R. (1989) ‘Human Resource Flexibility and Equity: Difficult Questions for Business, Labour and Government.’ Labour Research, 10(1), pp. 67–72.

238


Bertaux, D. (ed) (1981) Biography and Society, Sage, Beverly Hills. Blaikie, N. (1993) Approaches to Social Enquiry, Polity Press, Cambridge. Blatsou, C. (1996) A crop choice framework for a more sustainable agriculture: The case of the Argolid Valley in Greece, M.Phil Dissertation, Cranfield University. Boulding, K. (1968) ‘The Specialist with a Universal Mind.’ Management Science, 14(12), pp. B647–B653. Bowley, R., Young, T. and Colman, D. (1987) ‘A Systems Approach to Modelling Supply Equations in Agriculture’, Journal of Agricultural Economics, 38(2), pp. 151–166. Bozeman, B. (1986) ‘The Credibility of Policy Analysis: Between Method and Use.’ Policy Studies Journal, 144, pp. 519–539. Brown, J. (1987) ‘Assessing the Social Impacts of Hazardous Technology Facilities’, in Canter, D. (ed.) ‘Ethnoscapes; Transcultural Studies in Action and Place’. Bruntland Commission (1987) (World Commission on Environment and Development) ‘Our Common Future’, Oxford University Press, Oxford. Burgess, J., Limb, M. and Harrison, C.M. (1989) ‘Exploring Environmental Values through the medium of Small Groups’, Environment and Planning A, 20, pp. 309–320. Burgess, R.G. (1984) ‘In the Field: An introduction to Field Research’, George Allen and Unwin, London. Buttel, F. (1989) The Sociology of Agriculture: Current Conceptual Status, The Rural Sociologist, 9, pp. 16–26. Butzer, K. (1994) The Islamic traditions of Agroecology: Cross cultural experience, ideas and innovations, Ecumene, 1(1), pp. 7–50. Caldin, E.F. (1949) The Power and Limits of Science, Chapman & Hall, London. Calvin, E. (1949) The Power and Limits of Science, Chapman & Hall, London. Calvo-Cases, A., Soriano Soto, M., Boix Fayos, C., Cerdá Bolinches, A., Tiemessen, I. and Imeson, A. (1995) “Report of the Alicante study areas” A contribution to the ERMES (Environmental Response of Mediterranean Systems) Project EV5V-0023 as part of the EU-DGXII Environment Programme, Area IV/ 3, Desertification in the Mediterranean area. Carr, S. and Tait, J. (1991) Farmer’s Attitudes to Conservation, Built Environment, 16, pp. 218–230. Carrol, L. (1988) Sylvie and Bruno, Dover, Colchester. Cassell, J. (1988) ‘The relationship of observer to observed when studying up’, in Burgess, R.G. Studies in Qualitative Methodology, London, JAI Press. Caswell, M. and Zilberman, D. (1985) ‘The Choice of Irrigation Technologies in California’, American Journal of Agricultural Economics, 67(2), pp. 224–234. Checkland, P. (1981) ‘Systems Thinking, Systems Practice’, John Wiley, Chichester. Churchman, C.West et al. (1957) Introduction to Operations Research, Wiley, New York. Clark, N., Perez-Trejo, F. and Allen, P. (1995) Evolutionary Dynamics and Sustainable Development, Edward Elgar, Aldershot. Clayton, M. and Radcliffe, N. (1996) Sustainability: A Systems Approach, Earthscan, London. Cleveland, H. (1988) ‘Thesis of a new Reformation: The Social Fallout of Science 300 Years After Newton.’ Public Administration Review, 483, pp. 681–686. Clunies-Ross, T. and Hildyard, N. (1992) The Politics of Industrial Agriculture, The Ecologist, 22, pp. 65–71. Conway, G. (1985) Agroecosystem analysis, Agricultural Administration, 20, pp. 31–55. Cushman, D.P. and McPhee, R.D. (eds) (1980) “Message, attitude, behaviour relationship” Academic Press, London. Damianos, D., Damoussis, M. and Kasimis, C. (1991) ‘The Empirical Dimension of Multiple Job-Holding Agriculture in Greece’, Sociologica Ruralis, 31(1), pp. 37–47. Daneke, G. (1989) ‘On Paradigmatic Progress in Public Policy and Administration.’ Policy Studies Journal, 17 (2), pp. 277–296. de Vaus, D. (1996) ‘Surveys in Social Research’, Allen and Unwin, London, de Waal, C. (1991) ‘Urban Aspirations and Rural Development in Southern Greece’, PhD thesis, Cambridge University.

REFERENCES

239

Demoussis, M. and Sarris, A. (1988) ‘Greek Experience under the CAP: Lessons and Outlook’, European Review of Agricultural Economics, 15, pp. 89–107. Dorfman, J.H. and Heien, D. (1989) ‘The Effects of Uncertainty and Adjustment Costs on Investment in the Almond Industry’, Review of Economics and Statistics, 66, pp. 263–274. Dunn, W. (1981) ‘Public Policy Analysis.’ Prentice-Hall, Englewood Cliffs, N.J. Eckaus, R. (1987) ‘Appropriate Technology: The movement has only a few clothes on’, Issues in Science and Technology, Winter. Eden, C. (1992) On the nature of Cognitive Maps, Journal of Management Studies, 29(3). EEA (European Environment Agency) (1997) “Natural Resources CD-ROM” CORINE Land-Cover data, EUDGXI. Efstratoglou-Todoulou, S. (1988) ‘Current Experiences in Multiple Job-Holding in Greek Farming’, in Arkleton Trust, ‘Rural Change in Europe’. EIU Country Report (Greece), No. 1, 1992. ERMES annual report (1997) First Year Report of the ERMES project presented to DGXII of the European Union. Fanariotu, I. and Skuras, D. (1991) ‘Fertilizer Use and Environmental Policy in Agriculture: A Socio-economic Study in Greece’, Environmental Conservation, 18(2), pp. 137–142. Fielding, N. (1993) Qualitative Interviewing, in Gilbert, N. ‘Researching Social Life’ Sage, London. Fishbein, M. (1967) ‘Readings in Attitude Theory and Measurement’, Wiley, New York. Fishbein, M. and Ajzen, I. (1975) ‘Belief, Attitude, Intention and Behaviour’, Addison-Wesley, Reading M.A. Foddy, W. (1994) Constructing questions for interviews and questionnaires: Theory and practice in social research, Cambridge University Press, Cambridge. Food and Agriculture Organization (1985) Guidelines: Land Evaluation for Irrigated Agriculture. Soils Bulletin 55, Rome: FAO. Food and Agriculture Organization (1987) Yield Response to Water 33 Citrus Fruits, Rome: FAO. Forbes, H. (1976) We have a little bit of everything: The ecological basis of some agricultural practices in Methana, Trizinia, Annals of the New York acadamy of Sciences, 268, pp. 236–250. Fortune, J. and Peters, G. (1990) The formal systems paradigm for studying failures, Technology Analysis & Strategic Management, 2(4), pp. 383–390. French, B.C. and Bressler, R.G. (1962) ‘The Lemon Cycle’, Journal of Farm Economics 44, pp. 1021–1036. French, B.C. and Matthews, J.L. (1971) ‘A Supply Response Model for Perennial Crops’, American Journal of Agricultural Economics, 53, pp. 478–490. Fresco, L. and Kroonenberg, S. (1992) ‘Time A. and Spatial Scales in Ecological Sustainability, Land Use Policy, 9(3), pp. 155–168. Giddens, A. (1991) Structuration theory: past, present and future in Bryant, C. and Jary, D. (eds) Gidden’s theory of Structuration, Routledge, 1991. Gilbert, N. (1993) ‘Researching Social Life’ Sage, London. Glaeser, B. (1988) A Holistic Human Ecology Approach to Sustainable Agricultural Development, Futures, pp. 671–678. Granovetter, M. (1973) ‘The Strength of Weak Ties’, Amerxican Journal of Sociology, 78, pp. 1360–1380. Green, S. and Lemon, M. (1996) ‘Perceptual Landscapes in Agrarian systems: Degradation processes in Northwestern Epirus and the Argolid Valley, Greece’, Ecumene, 3(2), pp. 181–199. Harwood, R. (1989) Agroforestry and mixed farming systems, in Lugo, J. et al. (eds) Ecological Development in the Humid Tropics: Guidelines for Planners, Winrock International, Morrilton, pp. 271–301. Hassan, R.M. and Hallam, A. (1990) ‘Stochastic Technology in a Programming Framework: A Generalised Mean-Variance Farm Model’, Journal of Agricultural Economics, 41, pp. 158–179. Heady, E.O. (1952) Economics of Agricultural Production and Resource Use, Prentice Hall, Englewood Cliffs. Hedge, D. and Mok, J. (1987) ‘The Nature of Policy Studies: A Content Analysis of Policy Journal Articles.’ Policy Studies Journal, 16(1), pp. 49–62.

240


Heisey, P.W. et al. (1993), ‘Varietal Change in Post-Green Agriculture: Empirical Evidence from Wheat in Pakistan’, Journal of Agricultural Economics, 44, pp. 428–442. Herzfeld, M. (1992) ‘The Social Production of Indifference: Exploring the Symbolic Roots of Western Bureaucracy’, Berg, Oxford. Hofstede, G. (1995) Multilevel research of human systems: Flowers, bouquets and gardens, Human Systems Management, 14, pp. 207–217. Holling, C. (ed) (1978) ‘Adaptive Environmental Assessment and Management.’ John Wiley & Sons, New York. Horner, G.L. et al. (1992) The Canadian Regional Agricultural Model: Structure, Operation and Development, Technical Report 1/92 Agriculture Canada, Ottawa. Hornsby-Smith (1993) Gaining Access in Gilbert, N. ‘Researching Social Life’ pp. 52–67, Sage, London. House, R.M. (1987) USMP Regional Agricultural National Economic Division Report, Economic Research Service, USDA, Washington DC. Howitt, R.E. (1995) ‘A Calibration Method for Agricultural Economic Production Models’, Journal of Agricultural Economics, 46(2). Ibery, B. and Bowler, I. (1993) The Farm Diversification Grant Scheme, Adoption and nonadoption in England and Wales, Environment and Planning C: Government and Policy, 11, pp. 161–170. Imeson, A.C., Perez-Trejo, F. and Cammeraat, L.H. (1996) “The Response of Landscape Units to Desertification” in Brandt, C.J. and Thornes, J.B. (eds) “Mediterranean Desertification and Land Use.” John Wiley & Sons, London. Jantsch, E. (1967) ‘Technological Forecasting in Perspective.’ OECD. New York. Jeffrey, P. and Seaton, R. (1995) ‘The use of. Operational Research tools: a survey of Operational Research practitioners in the UK’, Journal of the Operational Research Society, 46, pp. 797–808. Jeffrey, P. (1996) Evolutionary analogies and Sustainability: Putting a human face on survival, Futures, 28(2), pp. 173–186. Jeffrey, P. (1997) A study of cross-disciplinary interaction, in Winder, N. and Van der Leuuw, S. (eds) Environmental Perception and Policy Making: Cultural and Natural Heritage and the Preservation of Degradation-Sensitive Environments in Southern Europe. Draft final report on contract EV5V-CT94– 0486 (Volume 1). Jeffrey, P. and Lemon, M. (1996) Understanding the Dynamics of Sustainable Communities: Stochasts, Cartesians and Social Networks, in O’Conner, M. and Ganslasser, U. (eds) ESE Conference Papers: Life Science Dimensions, Filander Press. Jeffrey, P. and Lemon, M. (1996) Understanding the Dynamics of Sustainable Communities: Stochasts, Cartesians and Social Networks, in O’Conner, M. and Ganslasser, U. (eds) ESE Conference Papers: Life Science Dimensions, Filander Press. Just, R. and Zilberman, D. (1986). ‘Does the Law of Supply Hold under Uncertainty?’, Economic Journal, 96, pp. 514–524. Kelly, G. (1955) The Psychology of Personal Constructs: A theory of personality, Norton, New York. Knippendorf, K. (1980) ‘Content Analysis: An Introduction to its Methodology’, Sage, Beverly Hills. Kraker, E. and Paddock, B. (1985) A Systems Approach to Estimating Prairie Crop Acreage, Working Paper, 15/85, Agriculture Canada Ottawa. Laboratory of Agricultural Hydraulics, Athens Agricultural University (1993) ‘The Problem of Argolis’, Archaeomedes First Year Report. Langer, E. (1975) ‘The Illusion of Control.’ Journal of Personality and Social Psychology, 32, pp. 311–328. Langer, E. and Roth, J. (1975) ‘The Effect of Sequence of Outcomes in a Chance Task on the Illusion of Control’, Journal of Personality and Social Psychology, 32, pp. 951–955. Larwood, L. and Whittaker, W. (1977) ‘Managerial Myopia: Self Serving Biases in Organizational Planning.’ Journal of Applied Psychology, 67, pp. 194–198.

REFERENCES

241

Lemon, M. and Jeffrey, P. (1998) Pathways elicitation I and II: Developing a process oriented approach to mapping personal evaluations of change, under review. Lemon, M. (1991) Perceptual Congruence and change: Non Urban Communities and Land-Use Planning, PhD. Thesis, Cranfield University. Lemon, M. and Longhurst, P. (1996) Towards a Conceptual Framework for Environmental Education, Proceedings of 3rd International Conference on Environmental Impact Assessment, 3, pp. 613–621. Lemon, M. and Naeem, A. (1990) ‘The Response to Non-Urban Change’, in Vyakarnam, S. (ed) ‘When the Harvest is in’, Intermediate Technology Publications, London. Lemon, M. and Park, J. (1993) ‘Elicitation of farming agendas in a complex environment’, Journal of Rural Studies, 9(4), pp. 405–410. Lemon, M. and Scamans, J. (1997) Individual attributes and adaptive capability: A case study of the CLA, in Armistead, C. and Kiely, J. (eds) Effective Organisations: Looking to the future, Cassel. Lemon, M., Hart, P. and Seaton, R. (1992) ‘Policy diversity in Europe: Consumer Perceptions and Local Government (The English Dimension)’, paper presented to XXIInd International Congress of Administrative Sciences, Vienna. Lemon, M., Seaton, R. and Park, J. (1994) ‘Social enquiry and the measurement of natural phenomena: the degradation of irrigation water in the Argolid Plain, Greece’, International Journal of Sustainable Development and World Ecology, 2(3), pp. 206–220. Lemon, M., Green, S. and Filippucci, P. (1997) Towards a comparative and collaborative framework, Final report to the EU (DGXII) on Environmental Perception Project (EV5V-0021). Lemon, M., Seaton, R. and Van de Leeuw, S. (1993) ‘Agricultural Policy and Desertification: A case-study of the Argolid, in Fantechi, R. and Luxembourg, R. (eds) Desertification in a rural context, European Commission. Lianos, T. and Parliarou, D. (1986) ‘Farm Size Structure in Greek Agriculture’, European Review of Agricultural Economics, 13, pp. 233–248. Linstone, H. (1981) The Multiple Perception Concept: With applications to Technology Assessment and other Decision Areas, Technological Forecasting and Social Change, 20, pp. 275–235. Linstone, H. (1981) The Multiple Perspectives Concept: With Applications to Technology Assessment and other Decision Areas, Technological Forecasting and Social Change, 20, pp. 275–325. Liverman, D. et al. (1988) ‘Global sustainability: Towards measurement’, Environmental Management, 12(2), pp. 133–143. Lockeretz, W. (1991) ‘Multidisciplinary research and sustainable agriculture’, Biological Agriculture and Horticulture, 8, pp. 101–122. Lowrance, R. et al. (1986) A Hierarchical Approach to Sustainable Agriculture, American Journal of American Agriculture, pp. 169–173. Madu, C. and Jacob, R. (1991) Multiple Perspectives and Cognitive Mapping to Technology Transfer Decisions, Futures, November. Manitea-Tsapatsaris, V. (1986) ‘Crisis in Greek Agriculture: Diagnosis and an Alternative Strategy’ , Capital and Class, 27, pp. 107–130. Mannermaa, M. (1991) ‘In Search of an Evolutionary Paradigm for Futures Research’, Futures, May, pp. 349– 372. Mannheim, K. (1950) Freedom, Power and Democratic Planning, Oxford University Press, New York. McLain, R. and Lee, R. (1996) Adaptive Management: Promises and Pitfalls, Environmental Management, 20 (4), pp. 437–448. Mihram, D. and Mihram, G.A. (1974) ‘Human knowledge: The role of models, metaphors and analogy’, International Journal of General Systems’, 1, pp. 41–60. Mikkelson, B. (1995) Methods for Development Work and Research: A Guide for Practitioners, Sage, London. Miller, G. (1980) Afterword’, in Message, Attitude, Behaviour Relationship, in Cushman, D. and McPhee, R. (eds) pp. 319–328, London, Academic Press.

242


Mlliken, F. (1987) ‘Three Types of Perceived Uncertainty About the Environment: State, Effect, and Response Uncertainty’. Academy of Management Review, 12(1), pp. 133–143. Molnar, J., Duffy, P., Cummins, A. and Van Santen, E. (1992) Agricultural Science and Agricultural Countercultures: Paradigms in search of a future, Rural Sociology, 57(1), pp. 83–91. Moore, M., Gollehon, N. and Carey, M. (1994) ‘Multicrop Production Decisions in Western Irrigated Agriculture: The Role of Water Price’, American Journal of Agricultural Economics, 76, pp. 859–874. Moore, M.R. and Negri, D.H. (1992) ‘A Multicrop Production Model of Irrigated Agriculture, applied to Water Allocation Policy of the Bureau of Reclamation’, Journal of Agricultural and Research Economics, 17(2), pp. 29–43. Moser, C.A. and Kalton, C. (1979) ‘Survey Methods in Social Investigation’, Heinemann, London. Murdoch, J. (1995) “Actor networks and the evolution of economic forms: combining description and explanation in theories of regulation, flexible specialization, and networks”, Environment & Planning A, 27, pp. 731–757. Murdoch, J. (1993) ‘Sustainable Rural Development: towards a Research Agenda’, Geoforum, 24(3), pp. 225– 241. Negri, D.H. and Brooks, D.H. (1990) ‘The determinants of Irrigation Technology Choice’, Western Journal of Agricultural Economics, 15, pp. 213–223. Nelson, R. and Winter, S. (1982) ‘An Evolutionary Theory of Economic Change.’ The Belknap Press of Harvard University, Harvard USA. Nerlove, M. (1989) The Dynamics of Supply: Estimation of Farmers Response to Prices, Baltimore, John Hopkins University Press. Newby, H. (1992) ‘Join forces in a modern marriage’, Time Higher Educational Supplement, 17th Jan, pp. 20. Newell, R. (1993) Questionnaires in Gilbert, N. ‘Researching Social Life’ Sage, London, pp. 94–115. Newson, M. (1992) ‘Land, water and development: River Basins and their Sustainable Management’, Routledge, London. Nieswiadomy, M. (1988) ‘Input substitution in Irrigated Agriculture in the High Plains of Texas’, Western Journal of Agricultural Economics, 13, pp. 63–70. O’Riordan, T. (1973) “Some reflections on environmental attitudes and environmental behaviour” Area, 5, pp. 17–21. O’Riordan, T. (1985) ‘What does sustainability really mean?’ in ‘Sustainable development in an industrial economy’, CEED Report. Ogg, C. and Gollehon, N. (1989) ‘Western Irrigation Response to Pumping Costs: A Water Demand Analysis Using Climatic Regions,’ Water Resources Research, 25, pp. 767–773. Okali, C. et al. (1994) Farmer Participatory Research, Intermediate Technology Publications, London. Open Systems Group (1983) Systems Behaviour, Open University, Milton Keynes. Oxley, T. and Mata Porras, M. “The emergence of socio-natural response units for the simulation of farmer decision making.” IERC, Cranfield University, (forthcoming) Oxley. T. (1994) “A Complex Systems Approach to Modelling Environmental Catastrophe.” PhD Thesis, IERC, Cranfield University. Pannell, D.J. (1987) ‘Crop Livestock Interactions and Rotation Selection’, in Kingswell R.S. and Pannell, D.J. (eds) CH. 4, MIDAS A Bioeconomic Model of a Dryland Farming System, Pudoc Wageningen, The Netherlands. Papaioannou, I. (1980) The water problem of the Argolid (in Greek), Agnessi Press. Park, J. (1993) ‘Towards a sustainable systems framework: The assessment of Silvoarable Agroforestry as an innovative cropping process’, PhD Thesis, Cranfield University . Park, J. and Seaton, R. (1995) ‘Research towards a sustainable systems framework: an integrative approach’, Agricultural Systems, 50, pp. 81–100. Pearce, D.W. and Turner, R.K. (1990) Economics of natural resources and the environment, Harvester Wheatsheaf.

REFERENCES

243

Perkins, H.C. (1988) ‘Bulldozers in the Southern Part of Heaven: Defending Place Against Rapid Growth’, Environment and Planning A, 20, pp. 285–308. Perry, G.M. (1986) Toward a Holistic Approach to the Cropping Mix Decision, PhD, Ann Arbor, Michigan. Polopolus, L. (1989) ‘Greek Agriculture in a Period of Adjustment’, Agriculture and Human Values, WinterSpring, pp. 82–90. Potter, C. and Gasson, R. (1988) ‘Farmer participation in voluntary land diversion schemes: Some predictions from a survey’, Journal of Rural Studies, 4(4), pp. 365–375. Prigogine, I. (1985) ‘New Perspectives on Complexity.’ in ‘The Science and Praxis of Complexity.’ United Nations University. Prigogine, I. (1984) ‘Order out of Chaos: Man’s New Dialogue with Nature’, Bantam Books, New York. Rosenbleuth, A. and Wiener, N. (1945) ‘The role of models in science.’ Philosophy of Science, 12, pp. 316–321. Rossini, F. and Porter, A. (1984) ‘Interdisciplinary Research: Performance and Policy Issues’ in Jurkovich, R. and Paelinck, J. (eds) ‘Problems in Interdisciplinary Studies.’ Gower, Aldershot. Ruiz, M. (1988) ‘Development of Mediterranean agriculture: An ecological approach’, paper presented to ‘Changing agriculture, changing landscape’, International Conference on the Future of the European Landscape, Rotterdam, 9–11 May. Sachdeva, P. (1984) ‘Development Planning: An Adaptive Approach’, Long Range Planning, 17(5), pp. 96–102. Sakellis, M. (1986) ‘An Empirical Analysis of Savings Behaviour of the Greek Agricultural Households’, European Review of Agricultural Economics, 13, pp. 169–188. Schaible, G.D., Kim, C.S. and Whittlesey (1991) ‘Water Conservation Potential from Irrigation Technology Transitions in the Pacific Northwest’, Western Journal of Agricultural Economics, 16, pp. 194–206. Schneider, J. Stevens, N. and Tornatzky, L. (1982) ‘Policy Research and Analysis: An Empirical Profile 1975– 1980’, Policy Sciences, 15, pp. 99–114. Seaton, R. and Cordey-Hayes, M. (1993) ‘The development and application of interactive models of industrial technology transfer’, Technovation, 13(1), pp. 45–53. Seaton, R. and Black, I. (1987) ‘Decision-support Systems and the Strategic Management of Train Service Punctuality’ in Murthy, T., Lawrence, L. and Rivier, R. (eds) ‘Computers in Railway Management’ Computational Mechanics Publications, Springer-Verlag. Sideris, A. (1934) Ai Georgikai Politikai tis Ellados kata tin Lixasan Ekatontaetian, Papadoyiannis, Athens. Silverman, D. (1993) ‘Interpreting Qualitative Data’, Sage, London. Simon, H. (1962) ‘The Architecture of Complexity’ Proceedings of the American Philosophical Society, 106(6), pp. 467–482. Slocombe, D. (1990) Assessing transformation and sustainability in the Great Lakes Basin, GeoJournal, 21(3), pp. 251–272. Spanopoulou, K. (1990) ‘Structural Weaknesses of Greek Agriculture in the Period 1950–1987’, Journal of Regional Policy, 10, pp. 519–554. Spedding, C. (1979) An Introduction to Agricultural Systems, Elsevier Applied Science. Spedding, C. (1991) ‘Thinking about the Future’, Journal of the RASE, 152, pp. 31–55. Springer, C., Herlihy, R. and Beggs, R. (1965) Advanced methods and models, Irwin, Homewood, Illinois. Stewart, F. (1987) ‘The Case for Appropriate Technology: A reply to R.S.Eckaus’, Issues in Science and Technology, Summer. Stubbs, M. and Lemon, M. (1996) Creating environmental information networks with a sense of audience: Some early lessons, Business strategy and the environment conference proceedings, ERP Environment, West Yorks, pp. 234–239. Stubbs, M. and Lemon, M. (1997) Injecting Creativity into Environmental Management: Lessons from a Dialogue on the UK National Air Quality Strategy, Business strategy and the environment conference proceedings, ERP Environment, West Yorks, pp. 232–237. Swan, J. (1995) Exploring Knowledge and Cognitions in Decisions about Technological Innovation: Mapping Managerial Cognitions, Human Relations, 48(11).

244


Tintner, G. (1941) ‘The Theory of Choice Under Subjective Risk and Uncertainty.’ Econometrica, 9, pp. 298– 304. Towriss, J.G. (1984) ‘A New Approach to the Use of Expectancy Value Models’, Journal of the Market Research Society, 26(1), pp. 63–75. Trott, P., Seaton, R. and Cordey-Hayes. M, ‘Inward Technology Transfer as an interactive process: A case-study of ICI’ Forthcoming in Technovation. Umali, D.L. (1994) Irrigation-Induced Salinity. A growing Problem for Development and the Environment, World Bank Technical Paper, No. 215, Washington DC. Urry, J. (1987) Society, Space, and Locality, Environment and Planning D: Society and Space, 5, pp. 435–444. Van Foerster, H. (1977) ‘The Curious Behaviour of Complex Systems: Lessons in Biology’, in Linstone, H.A. and Simmonds, W.H.C. (eds) ‘Futures Research: New Directions’, Addison-Wesley, Reading. Van Tuijl, W. (1993) ‘Improving Water Use in Agriculture: Experiences in the Middle East and North Africa’, World Bank Technical Paper, No. 201. von Bertalanffy, L. (1983) General Systems Theory—A Critical Review, in Open Systems Group Systems Behaviour, Open University, Milton Keynes, pp. 59–79. Wallace, A. (1976) ‘Influence of Salty Soils on Citrus Trees in Greece’, Project (76)05, OECD, Technical Cooperation Service, CT/1180. Walters, C. (1986) ‘Adaptive Management of Renewable Resources.’ Macmilan. New York. Weaver, W. (1948) ‘Science and Complexity.’ American Scientist, October, pp. 536–544. Whyte, W.F. (1955) ‘Street Corner Society’, University of Chicago Press, Chicago. Wilson, B. (1984) Systems: Concepts, Methodologies and Applications, Wiley, Dorchester. Winder, N. and Van der Leuuw, S. (eds) (1997) Environmental Perception and Policy Making: Cultural and Natural Heritage and the Preservation of Degradation-Sensitive Environments in Southern Europe. Draft final report on contract EV5V-CT94–0486 (Volumes 1–4). Winder, N., Jeffrey, P. and Lemon, M. (1998) “Nested Master-Equation models: An application to the simulation of crop choice dynamics.” Etudes et Recherches sur les Systemes Agraires et le Developpement’, (Forthcoming). Winter, S. (1964) ‘Economic Natural Selection and the Theory of the Firm.’ Yale Economic Essays, 4, pp. 224– 272. Yin, R.K. (1989) ‘Case Study Research’, Sage, London.

INDEX

access in social enquiry 45 see also social enquiry, interviews, elicitation actors 2, 3 adaptability (adaptive capacity) 2, 5, 15, 71, 232 agencies 3, 21 Agricultural University of Athens 183, 196, 200 agricultural efficiency 15, 25, 133 see also structural weaknesses agricultural extension 151 agricultural production 20, 79 see also cropping, crop characteristics, markets air mixers 129 see also frost analogies (biological, evolutionary) 18, 75 Anavalos spring water 10, 34, 37, 103, 122 see also canals, infrastructure aquifers 116, 165, 186 archaeological sites (Mykenes, Tiryns, Argos, Epidavros) 28, 34, 140 ARCHAEOMEDES (Understanding the natural and anthropogenic causes of land degradation and desertification in the Mediterranean Basin) 31 Argolid Valley (Peloponnese) 25 Argos 34, 36, 82 artificial recharge 36, 122, 124 see also water depletion, flooding Association of Argolid Agronomists 119 attitudes 40 to farm work 145 to public administration 150 to ecological agriculture 21 and behaviour 40

cost of drilling 120 see also zones boundaries disciplinary 13, 15, 37 administrative 30 systems 68, 71 Bruntland Report 15

biomass 222 see also models and vegetation bore-holes and irrigation pumps 116 permits for 119

dealers 108 see also markets decision making 4 opportunity space 40, 76, 162, 232

canals 35 Argolid 116 Marina Baixa 219 climate 65 climate change 212 Argolid 31 see also frost, flooding co-operatives 107 see also markets complexity x, 68 connectivity 72 crop production 18, 79 mono-cropping 2, 25, 81 multi-cropping 36, 81 farmer typology and cropping 162 crop characteristicsapricots 82; vines 85; olives 87; tobacco 89; vegetables 89; lemons 92; oranges 93 crop choice model (framework) 165

245

246

INDEX

decision space 40, 162, 232 DEH (Public Electric Company) 119 degradation 2, 10, 33 see also water depletion, salination depopulation 29, 89, 136, 146, 148 desertification response units 210, 219 diversity 2, 15, 18, 27, 71, 162, 226 see also multi-cropping, multiple job holding efthinofovia (fear of responsibility) 149 elicitation 40 cognitive mapping 40 see also ethnography, interviews emergence 72 equilibrium 72 ERMES (Environmental Response of Mediterranean Systems) 212, 214 ethnography 41 European Union on efficient farming structures 145 and citrus restructuring 93, 107, 112 concerns about 46 see also crop characteristics, farm size, multiple job holding evolutionary theory 15 see also diversity, resilience, adaptability

inputs 158, 207, 212 and private agronomists 152 integrative method 15 interviews (structure for field work) 61 question design and piloting 56 sampling 58 snowball sampling 41 coding and analysis 58 see also social enquiry integrated dynamic model 210 irrigation see Anavalos, bore-holes, water depletion, sprinkler systems, free flow irrigation isolation techniques 118 see also salination juicing 111 knowledge in social enquiry 48 Land Improvement Service (YEB) 119, 128

geomorphology 226 Greek mythology and water 34

Marina Baixa—Alicante, Spain 210 see also integrated dynamic model markets 105 external markets 112 local markets 108 see also dealers Medlar 229 see also Marina Baixa messy problems 9 migration (depopulation) 29, 87, 136 see also migrant labour migrant (seasonal) labour 35, 87, 97, 108, 147 models and modelling 72 abstract models 74 typology of models 75 of vegetation 219 interpretative lenses 232 see also crop choice, strategic and integrated dynamic models multiple job holding (pluriactivity) 142

hierarchy 72

Nafplio 29, 36

information 72 storing information 148 infrastructure 2, 6, 10, 31 see also Anavalos

perception of uncertainty in farming 156 policy relevant research x, 226 public administration and bureaucracy 148 and political patronage 150

farmer characteristics 146 age and gender 146 status of farming—education of farmers 147 typology of farmers in the Argolid Valley 162 farm size, fragmentation and ownership 137 feedback 8, 72 free flow (flood irrigation) 124 flooding 31, 34, 64, 116, 127 see also climate, artificial recharge former Yugoslavia 113 frost 128 see also air mixers, sprinkler systems—spraying

INDEX

see also agricultural extension and information requisite variety 72 resilience 6, 15, 71 Rhone Valley 212 salination 67, 175, 194 data collection 186 salt movement equations 204 effect on citrus trees 100 Sharka virus 37, 84 see also apricots simulation 222 social enquiry 37 and the measurement of natural phenomena 43 levels of 45 choice of technique 45 see also ethnography, elicitation soils in the Argolid 30 sprinkler systems for irrigation and frost protection 124 see also frost strategic model of water flow 197 stakeholders 2 see also actors, agencies structural weaknesses in Greek agriculture 133 see also farm size and ownership, farmer characteristics subsidiarity 2 sustainability 15 sustainable pathways 19 systems thinking 70 see also complexity, feedback, uncertainty, messy problems technology 113 see also infrastructure, Anavalos, bore-holes, air mixers etc. topography (slope, aspect) 219 tourism 29 agrotourism 154 triangulation (multiple techniques) 24 water depletion 183 effect on citrus trees 100 see also salination zones (procedure for establishing) 100 and crop production 101, 178

and bore-hole depth and salination 191 and irrigation technologies 124 and farm size 140

247

Exploring Environmental Change Using an Integrative Method

Phenology: An Integrative Environmental Science

Phenology: An Integrative Environmental Science

Temperature and Toxicology: An Integrative, Comparative, and Environmental Approach

Exploring Environmental History: Selected Essays

Environmental Modeling Using MATLAB

Timescales and Environmental Change

Climate Change (Environmental Issues)

Global Environmental Change

Psychology and Environmental Change

Global Environmental Change

Glaciers and environmental change

Climate Change (Environmental Issues)

Microarrays for an integrative genomics

Neuropsychiatric Disorders: An Integrative Approach

Environmental Change in South-East Asia (Global Environmental Change Series)

Natural hazards and environmental change

Encyclopedia of Global Environmental Change

Postmodern Climate Change (Environmental Politics)

Glacier Science and Environmental Change

Natural Hazards and Environmental Change

Geomorphology and Global Environmental Change

Titan: Exploring an Earthlike World

Depression and Women: An Integrative Treatment Approach

Soil and Water Chemistry: An Integrative Approach

Abnormal Psychology: An Integrative Approach, 6th Edition

The Psychology of Humor: An Integrative Approach

The finite element method using MATLAB

Data Analysis Using the Least-Squares Method

Cognitive Rehabilitation: An Integrative Neuropsychological Approach

Soil and water chemistry. An integrative approach

Exploring Environmental Change Using an Integrative Method

Phenology: An Integrative Environmental Science

Phenology: An Integrative Environmental Science

Temperature and Toxicology: An Integrative, Comparative, and Environmental Approach

Exploring Environmental History: Selected Essays

Environmental Modeling Using MATLAB

Timescales and Environmental Change

Climate Change (Environmental Issues)

Global Environmental Change

Psychology and Environmental Change

Global Environmental Change

Glaciers and environmental change

Climate Change (Environmental Issues)

Microarrays for an integrative genomics

Neuropsychiatric Disorders: An Integrative Approach

Environmental Change in South-East Asia (Global Environmental Change Series)

Natural hazards and environmental change

Encyclopedia of Global Environmental Change

Postmodern Climate Change (Environmental Politics)

Glacier Science and Environmental Change

Natural Hazards and Environmental Change

Geomorphology and Global Environmental Change

Titan: Exploring an Earthlike World

Depression and Women: An Integrative Treatment Approach

Soil and Water Chemistry: An Integrative Approach

Abnormal Psychology: An Integrative Approach, 6th Edition

The Psychology of Humor: An Integrative Approach

The finite element method using MATLAB

Data Analysis Using the Least-Squares Method

Cognitive Rehabilitation: An Integrative Neuropsychological Approach

Soil and water chemistry. An integrative approach

Recommend Documents