Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Fishery planning in Barbados: the implications of social strategies for coping with uncertainty McConney, Patrick A. 1995

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1995-060195.pdf [ 6.38MB ]
Metadata
JSON: 831-1.0088377.json
JSON-LD: 831-1.0088377-ld.json
RDF/XML (Pretty): 831-1.0088377-rdf.xml
RDF/JSON: 831-1.0088377-rdf.json
Turtle: 831-1.0088377-turtle.txt
N-Triples: 831-1.0088377-rdf-ntriples.txt
Original Record: 831-1.0088377-source.json
Full Text
831-1.0088377-fulltext.txt
Citation
831-1.0088377.ris

Full Text

FISHERY PLANNING IN BARBADOS: THE IMPLICATIONS OF SOCIAL STRATEGIES FOR COPING WITH UNCERTAINTY by PATRICK ADRIAN McCONNEY B.Sc., The University of Wales, 1981 M.E. S., Daihousie University, 1987 A THESIS SUBMITTED TN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES Interdisciplinary Studies (Resource Management Science)  We accept this thesis as conforming ? e required standard  THE UNIVERSITY OF BRITISH COLUMBIA  June 1995 © Patrick Adrian McConney, 1995  In presenting this thesis in partial fulfillment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted the head by of my department by or his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission.  (Signature)  MCQH  Department of  The University of British Columbia Vancouver, Canada  Date  2(,  Lqq  “L  11  ABSTRACT  This thesis provides information relevant to fisheries in which there is a desire to establish fisherfolk organizations, but where individualistic social networks rather than social cohesion and community prevail. Such situations in small-scale fisheries are poorly documented, but may define limits to the feasibility of co-management. In the case researched, the government of Barbados is designing a fisheries management planning process, but there is insufficient information on the social and economic characteristics of the unmanaged, small-scale, commercial fishery for migratory pelagics to determine whether either a state-structured (bioeconomic) or a cooperative (co-management) approach is appropriate. As a contribution to solving this practical, interdisciplinary problem, this study examined: the fishery-related uncertainties perceived by fisherfolk and government officials in Barbados; the social strategies of atomism, personal networks and formal organizations that fisherfolk may use to cope with uncertainty; and, whether the most appropriate initial management planning approach is bioeconomic or co-management. Research was conducted in Barbados between November 1993 and September 1994 involving surveys, social network analysis, participant observation and the study of official documents. Uncertainties related to fish catch and price were perceived by the majority of fisherfolk to be the most problematic, and the analysis focused on the means of coping with these. Evidence of social atomism was weak. Social networks, which tended to be individualistically-oriented among fishers, boat owners and processors, but more cooperative among vendors, were prevalent. Attempts by the harvest sector to formally organize to obtain market power had failed, but efforts to use this strategy persist. The state was found deficient in fishery planning and management capability. Barriers to communication within the state, and between it and the industry were apparent. Due mainly to the prevalence of networks and the state’s deficiencies, the bioeconomic approach is judged to be inappropriate in this setting. Due mainly to the high level of individualistic competition, the repeated failure of harvest sector organizations, and barriers to communication, co-management is problematic but more likely to be successffil. An incremental, institution-building approach to co-management is proposed due both to the flexibility of this approach and to the current political and planning environment that favours participative initiatives.  111  TABLE OF CONTENTS  ABSTRACT  ii  TABLE OF CONTENTS  iii  LIST OF TABLES  vii  LIST OFFIGURES  ix  ACKNOWLEDGMENT  x  1. OVERVIEW 1.1 BACKGROUND 1.2 RESEARCH CASE 1.3 ARRANGEMENT OF CHAPTERS  2. RESEARCH PROBLEM, QUESTIONS AND CONCEPTS 2.1 RESEARCH PROBLEM AND QUESTIONS  2.1.1 2.1.2 2.1.3 2.1.4 2.1.5  General problem Research problem Research questions Key concepts Relationships between concepts  3. LII’.KS TO THE LITERATURE 3.1 BIOECONOMICS VERSUS CO-MANAGEMENT  3.1.1 Bioeconomic model 3.1.2 Co-management 3.2 FISHERIES-RELATED UNCERTAINTY  3.2.1 Perceptions and planning 3.2.2 Uncertainty is ubiquitous 3.2.3 Uncertainty in the Caribbean context 3.3 SOCIAL STRATEGIES FOR COPING WITH UNCERTAINTY 3.3.1 Social atomism versus embeddedness 3.3.2 Social networks and organizations 3.3.3 Sampling Caribbean sociology 3.3.4 Social strategies in the fisheries literature 3.4 PERSPECTIVES ON PLANNING  3.4.1 Information as a social power resource 3.4.2 General considerations 3.4.3 Integrating social science into fishery planning 3.4.4 Co-management and participatory planning 3.4.5 Planning development in the Caribbean 3.4.6 Progress in fishery planning 3.4.7 Using this study as a planning resource 4. THE BARBADOS SITUATION  1 1 3 4 6 6 6 7 8 9 11 16 16 16 18 20 20 22 24 29 29 30 33 37 40 40 41 43 45 50 54 56 63  iv 4.1 COUNTRY PROFILE  4.1.1 Physical conditions 4.1.2 People and economy 4.2 THE FISHING INDUSTRY 4.2.1 Pelagic resources 4.2.2 Harvest sector 4.2.3 Postharvest and supporting activities 4.2.4 Administration and organizations 5. RESEARCH DESIGN AND METHODS 5.1 RESEARCH DESIGN 5. 1. 1 General considerations  5.1.2 Perceptions of uncertainty 5.1.3 Social strategies for coping 5.1.4 Implications for planning 5.2 RESEARCH METHODS  5.2.1 5.2.2 5.2.3 5.2.4 5.2.5 5.2.6  Survey of fishing industry Social network measurement Survey of government officers Other interviews Participant observation Document study 5.2.7 Discussion 6. FISHERFOLK AND UNCERTAINTY IN THE PELAGIC FISHERY 6.1 SURVEY RESPONDENTS  6.1. 1 Occupational categories 6.1.2 Respondent profile 6.2 PERCEPTIONS OF UNCERTAINTY  6.2.1 General uncertainty 6.2,2 Specific uncertainties 6.2.3 Major uncertainties 6.3 OFFICIAL AND SCIENTIFIC MEASUREMENTS OF UNCERTAINTY 6.3.1 Fish catches  6.3.2 Fish prices 6.3.3 Other uncertainties 6.3.4 A note on uncertainty reduction through information exchange 6.4 CONcLuDII’JG DISCUSSION  6.4.1 Summary 7. SOCIAL STRATEGIES FOR COPING WITH UNCERTAINTY: ATOMISM AND NETWORKS 7.1 SURVEY RESULTS 7.1.1 Income from the fishing industry 7. 1.2 Occupational alternatives and credit 7.1.3 Kinship 7.1.4 Degree of fisherfolk interaction 7.1.5 Social network ties and contents 7.1.6Power 7.1.7 Desired industry improvements  .63 63 63 65 65 69 74 78 83 83 83 84 85 85 86 86 94 98 100 102 103 104 106 106 106 108 112 113 113 121 122 122  130 134 136 137  139 141 141 141 143 145 146 146 150 151  V  .153 155 160 161 175 184 191 195 202  7.2 SoCIAL NETWORK GENERAL FEATURES 7.2.1 Aid utilized 7.3 FISHING INDUSTRY NETWORK DETAILS 7.3.1 Fishers as focal individuals 7.3.2 Owners as focal individuals 7.3.3 Vendors as focal individuals 7.3.4 Processors as focal individuals 7.4 CONCLUDING DISCUSSION  7.4.1 Summary  8. SOCIAL STRATEGIES FOR COPING WITH UNCERTAINTY: ORGANIZATIONS .204 204 8.1 ATTITUDES TOWARDS ORGANIZATION 204 8.1.1 Fishers 205 8.1.2 Owners 205 8.1.3 Vendors 206 8.1.4 Processors 206 8.1.5 Summary 207 8.2 ORGANIZATIONAL ALTERNATIVES 208 8.2.1 Cooperatives 218 8.2.2 Union 222 8.2.3 Associations 223 8.2.4 Company 225 8.3 BUFFA: A MICRO-STUDY OF ORGANIZATION FORMATION 226 8.3.1 Origins and background 229 8.3.2 Membership 232 8.3.3 Management 234 8.3.4 Relationship with government 236 8.3.5 Current status 237 8.4 CONCLUDING DISCUSSION 244 8.4.1 Summary .  9. IMPLICATONS FOR FISHERY PLANNING 9.1 SURVEY RESULTS  9.1.1 9.1.2 9.1.3 9.1.4 9.1.5  Barriers to joint efforts Fisheries management planning Fisherfolk involvement in management planning Information for planning: usefulness and availability Information on uncertainties and strategies  9.2 ILLUSTRATIVE EXPERIENCES  9.2.1 Tent Bay infrastructure project 9.2.2 Fisheries legislation public consultation 9.2.3 Reads Bay Coastal Zone Management Project 9.2.4 Early Fisheries Advisory Committees 9.2.5 Bridgetown Fisheries Complex Advisory Committee 9.2.6 Fish Importation and Marketing Committee 9.2.7 Oistins boat trailer cooperation 9.2.8 National Development Plan 9.3 PRESENT PLANNING ENVIRONMENT  9.3.1 Legal framework  246 246 247 250 252 257 259 260 261 262 264 266 267 269 270 272 275 275  vi 9.3.2 Political aspects 9.3.3 Status of the Fisheries Division 9.4 CONCLUDING DISCUSSION 9.4.1 Summary  .277 280 283 294  10. CONCLUSION  296  11. BIBLIOGRAPHY  307  12. APPENDIX A: QFISH SURVEY  322  13. APPENDIX B: QGOVT SURVEY  327  14. APPENDIX C: SOCNET STUDY  332  15. APPENDIXD: SUPPLEMENTARY DATA  338  vii LIST OF TABLES  Table 4. iGross domestic product and employment in Barbados, 1992 Table 4.2 The recent history of the fishing industry in Barbados Table 4.3 Contribution of major pelagic species to estimated total landings Table 4.4 Boat ownership by gender, concentration and type Table 4.5 Estimated fishing income and expenses Table 4.6 Estimated total fish landings, contribution to GDP and trade Table 5.1 Harvest sector populations Table 5.2 Quota allocations and actual interviews executed Table 6.1 Distribution of fishers by boat type Table 6.2 Boat ownership among respondents Table 6.5 Government officers’ perceptions of contributors to uncertainty Table 6.7 Seasonality in the pelagic fishery at Oistins Table 7.1 Proportion of income from fishing industry Table 7.2 Fisherfolks alternative work in and out of the fishing season Table 7.3 Involvement of fisherfolk kin in the industry Table 7.4 Means of coping with fishing industry uncertainty Table 7.5 Catch information networks Table 7.6 Price information networks Table 7.7 Marketing networks Table 7.8 Small loan networks Table 7.9 Large loan networks Table 7.10 Most useful types of aid received Table 7.11 Sources of most useful types of aid received Table 7.12 Who has the most influence on how the fishing industry operates Table 7.13 Most important desired improvements to the fishing industry Table 7.14 Who should improve the fishing industry Table 7.15 Perception of difficulty in state/industry cooperation Table 7.16 Reasons for difficulty in state/industry cooperation Table 7.17 Respondents’ strategy for coping with uncertainty Table 7.18 Perceptions of others’ strategies for coping with uncertainty Table 7.19 Fisherfolk’s average network strands of aid Table 7.20 Fisherfolk’s average network composition by member category Table 7.21 Fisherfolk’s average network composition by ties, gender and affiliation Table 7.22 Multistranded ties (number of strands per tie) Table 7.23 Proportion of friends in the fishing industry Table 7.24 Linkage of network members Table 7.25 Fisherfolk’s average network by extent of reciprocal and conflictual ties Table 7.26 Aid received from the network of a fisher Table 7.27 Aid received from members in the network of an owner Table 7.28 Aid received from members in the network of a vendor Table 7.29 Aid received from members in the network of a processor Table 9.1 Reasons for difficulty in state/industry cooperation (copy of Table 7.16) Table 9.2 Acquaintance with the term “fisheries management plan” Table 9.3 Meaning of “fisheries management plan” Table 9.4 Preferred fisherfolk involvement in management planning Table 9.5 Usefulness of information for planning  64  66 67 71 72 78 88 90 107 108 120 126 142 144 145 146 148 148 148 149 149 149 150 150 152 153 153 153 155 155 156 156 158 158 158 159 160 161 175 184 191 247 250 250 252 257  viii Table 9.6 Availability of information for planning Table 9.7 Fisheries Division’s budget within the Ministry of Agriculture Table 9.8 Applicability of factors that favour comanagement  258 282  289  ix LIST OF FIGURES Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure  3.1 Social research framework 3.2 Ocean environment around Barbados 3.3 Social atomism, networks and formal organization 3.4 Fishery planning approaches 4.1 Fishing boats and methods 4.2 Fish landing sites 4.3 Simplified fish marketing and distribution channels 6.1 Uncertainties perceived by all fisherfolk 6.2 Uncertainties perceived by fishers 6.3 Uncertainties perceived by fishing and non-fishing owners 6.4 Uncertainties perceived by vendors and processors 6.5 Oistin’s total pelagic fish catches, 1981 to 1984 6.6(a) to (e) Seasonality of major pelagic species at Oistins, 1981 to 1994 6.7(a) to (e) Variation in fish prices for the major pelagic species at Oistins 9.1 Fisherfolks’ political opinion  21 25 31 58 70 75 76 114 115 118 119 124 124 131 279  x ACKNOWLEDGEMENT  I would like to thank the Government of Barbados for granting me the leave to conduct this research funded under the CARICOM Fisheries Resource Assessment and Management Program (CFRAMP). I am grateful for the invaluable guidance of my research Supervisory Committee: Professors L. Lavkulich, J.D. McPhail, B. Elliott, P. Boothroyd and N. Guppy (Research Supervisor). I particularly appreciate my Research Supervisor taking the time and interest to visit me in the field and experience the fishery firsthand. My greatest debt of gratitude, however, is owed to the very many people in the Barbados fishing industry who shared their thoughts and experiences with me, and without whose interest and cooperation this study would not have been possible. Among these, special thanks are due to the founders of the Barbados United Fisherfolk Association who allowed me the privilege of participating in the formation of the organization. The assistance of my colleagues in the Fisheries Division and the Ministry of Agriculture is also deeply appreciated. Last, but not least, I am very grateful to Pamela and Jamila, my wife and daughter, for their motivation and support throughout this project.  1 1. OVERViEW  The purpose of this thesis is to examine the possible implications of social relations and social organization in a fishing industry for planning the introduction of a state-structured (bioeconomic) or cooperative (co-management) approach to fishery management. The purpose is met through a case study of Barbados’ unmanaged tropical small-scale commercial fishery for migratory pelagics. Special attention is paid to the use of information on perceptions of uncertainty as a planning resource. This chapter provides an overview of the thesis. 1.1 BACKGROUND  In recent times, neoclassical economics has been combined with fisheries biology to become the dominant theoretical influence on fisheries management and planning (Wilson et a!. 1994). In the conventional model used for planning it is assumed that individuals in the fishing industry are socially atomistic and that the state should be responsible for over-seeing fishery management decision-making (Jentoft 1989). This bioeconomic model, based on the open access common property resource theme, ignores the empirical evidence from sociology and anthropology of the social embeddedness of economic action. Embeddedness is demonstrated, for example, in the strategies involving the social networks and organizations that some people use to manage fisheries and cope with the uncertainties of the fishing industry (Acheson 1981). Although the bioeconomic perspective continues to predominate, its preeminence is declining because of its inability to encourage sustainable fishing practices in real situations (Ludwig et a!. 1993; Wilson et al. 1994). Support is growing for co-operative management (co management) as an alternative (Pinkerton 1989; McGoodwin 1990; Dyer and McGoodwin 1994). Co-management is “power-sharing in the exercise of resource management between a government and a community or organization of stakeholders” (Pinkerton 1992:33 1). This  2  alternative approach rejects social atomism and accepts the social embeddedness of fishing as an economic activity. How universally co-management might be applied has recently been questioned. Kuperan and Abdullah (1994) suggest that there are limits to co-management where resource, political and socioeconomic factors hinder social cohesion or organization formation among fisherfolk. Associated with the bioeconomic and cooperative management models are characteristic approaches to the planning that precedes and accompanies fishery management. Planning associated with the bioeconomic model tends to be structured by the state, and dominated by biological and economic concerns, ignoring social factors due to the atomism assumption (Jentoft 1989). In contrast, co-management typically involves cooperative planning as an expression of power-sharing. It can potentially, but not necessarily, include a wide range of social issues in addition to biological and economic concerns (Pinkerton 1989). Cycon (1986) criticizes the unquestioning use of the neoclassical bioeconomic fishery model by international development agencies and beneficiary governments, claiming it unrealistically assumes social atomism, and is problematic on this account. Furthermore, he claims that international aid agencies and fisheries planning agencies in developing nations are not geared toward consideration of social issues, regardless of the approach attempted. This implies that, whether planning agencies seek to introduce either bioeconomic or cooperative management approaches, without applied socioeconomic research on social relations and organization they run the risk of using an unsuitable approach. Social scientists have long argued for the inclusion of situation-specific socioeconomic data in fishery planning (Cycon 1986; Orbach 1986; Wilson etal. 1994). In summary, there is no universally accepted approach to fishery planning or management. Depending on the situation, either the bioeconomic or co-management approach or perhaps a  3  new formulation, may be most suitable. Fisheries plans may fail due to the use of inappropriate socioeconomic conceptions of society and insufficient use of socioeconomic information (Poggie 1992). Integrating applied social science into fisheries planning processes could make them more proactive and better able to direct the collection and evaluation of socioeconomic data (Dyer 1994). Such data could provide information useful for designing the process for management planning and for subsequently executing the process and managing the fishery. 1.2 RESEARCH CASE  In the case that I have chosen to research, the government of Barbados is at the critical point of designing its fisheries management planning process with international assistance. The thesis research problem is that it is not clear, in terms of the social relations and organization in the pelagic fishery, whether it is most feasible to introduce a bioeconomic approach, a co management approach or neither. Without social information the appropriate approach to fisheries planning and the consequences of its use will be difficult to determine. The applied interdisciplinary research undertaken in Barbados from November 1993 to September 1994 involved the collection of pertinent socioeconomic data on the small-scale, commercial, open access fishery for migratory pelagics.  The choice of an applied  interdisciplinary approach with a social focus reflects, first, an enduring perspective in the Caribbean that, given the limited human and other resources of small island developing states, social research should be applied to real problems (Lindsay 1978). And second, that although the path towards sustainable development in the Caribbean is thought to be through interdisciplinary approaches, social issues have previously been neglected in planning (Cox and Embree 1990). Specifically, the study first examines the fishery-related uncertainties that people in the fishing industry and relevant state agencies perceive. Then it examines the social strategies that  4 people in the Barbados pelagic fishing industry use to cope with their perceived uncertainties. It evaluates whether their strategies are socially atomistic or embedded in personal networks or in formal organizations. It explores the implications of this information on perceived uncertainties and strategies for fisheries planning in Barbados, with emphasis on evaluating the most suitable planning approach. Finally, it indicates the significance of the Barbados case for the general literature on fisheries management planning. 1.3 ARRANGEMENT OF CHAPTERS  The next chapter reviews the research problem, questions, and concepts. Chapter Three describes the linkages between this research and existing literature, and Chapter Four provides details on the historical and contemporary situation in Barbados. Interviews were used to gather data on perceptions, social strategies and planning from the people involved in the fishing industry and from government officials. These data were supplemented by evidence from documents along with observation of, and participation in, the activities of the fishing industry. Chapter Five describes research design and methods. Chapter Six, the first results and discussion chapter, addresses the first research question. This is done by examining the perceptions people have of uncertainty, and how these relate to scientific and statistical data on the fishery. Chapters Seven and Eight address the second research question dealing with social strategies for coping with uncertainty. The focus is on the relative importance of, and problems with, social atomism, social networks and formal organizations as coping strategies based on the data from all sources. Chapter Nine deals with the last question: the implications of the research results for approaches to fisheries planning. It examines the relationship between fisheries management and development planning in Barbados and whether the co-management or the bioeconomic  5 approach is likely to be better suited to the Barbados situation. Chapter Ten reiterates the study’s major findings and contribution to knowledge. Following this chapter is the Bibliography and Appendices.  6 2.  RESEARCH PROBLEM, QUESTIONS AND CONCEPTS  2.1 RESEARCH PROBLEM AND QUESTIONS 2.1.1  General problem  We all face uncertainty in almost every endeavour, and we generally try to devise strategies for coping with uncertainty based on our measurement or perception of it (Lorenzi 1980). In the fisheries context it is generally accepted that uncertainty permeates our attempts to understand ecological systems (Ludwig et a!. 1993). These systems include marine pelagic fisheries in which uncertainty affects fisherfolk (fishers, boat owners, fish vendors and processors) and state officials (fishery scientists, managers and planners). Yet the fisheries management and planning literature focuses mainly on how officials measure and perceive ecological and economic uncertainty, and how the scientific and legal-institutional strategies they devise ultimately affect fisherfolk. For example, state fishery planners often set plan objectives, strategies and tactics taking measurements and their perceptions of uncertainty into account (MacKenzie 1974). But understanding what uncertainties fisherfolk perceive, and why they chose particular strategies to cope with them, may also be important for planning. I do not mean cognitive strategies such as ignoring uncertainty, but social strategies ranging from individual to collective and from informal to formal. For example, some uncertainties may be perceived to be of such magnitude that fisherfolk devise social coping strategies ranging from social atomism or individualism, through loose or closely knit personal networks, to formal organizations such as cooperatives (Polinac 1988a). These social aspects of the fishery have received less attention in the fisheries management and planning literature. Information on the uncertainties that people in the fishing industry perceive, and the social strategies they use for coping, can be a potentially powerful resource for fishery planning by  7 contributing a realistic social foundation to the planning model. However, as mentioned in the overview, the currently predominant bioeconomic approach to fishery planning typically assumes social atomism in the fishing industry and largely ignores social information (Poggie 1992). The increasingly favoured co-management approach typically requires, among other things, that social cohesion and the capability to form organizations exists in the fishing industry or can be induced, most commonly at the community level (Pinkerton 1989). Social networks are either omitted, or are seen as the causes of model imperfections from the bioeconomist’s perspective (Terkla et al. 1988). Networks are typically seen as possible precursors to greater organizational formality from the co-management perspective (Acheson 1989). Using the three categories of strategy—social atomism, social networks and formal organizations—several scenarios are conceivable, and a few are summarized here. It is possible, for example, that the fishing industry may exhibit more than one social strategy simultaneously, there being differences by occupation, gender, age or other variables. The mix of strategies or a single strategy may be dynamic, changing over time perhaps in response to uncertainty. Strategies may be conflicting—tendencies towards atomism and network obligations working against effective organization, for example. Since one requires an absence of social ties and the other strong social ties, neither the bioeconomic nor co-management models may be satisfied by the social situation in the fishery if it is intermediate or mixed. This can only be discovered through empirical investigation. 2.1.2 Research problem  In the case that I have chosen to research, the government of Barbados has stated, and put into law, its intention to plan the management and development of its fishing industry. The major fishery is a seasonal, small-scale, commercial, open access fishery for shared stocks of migratory pelagics. My observations as a fishery officer in Barbados suggest that, while this fishery seems  8  individualistic, it is not atomistic. Social relations both enable and constrain fishery activities, but broad social cohesion and industry-wide formal organizations are weak. The research problem is that it is not clear, in terms of the social relations and organization in the pelagic fishery, whether it is most feasible to introduce a bioeconomic approach, a co management approach or neither. There is no documented information on perceptions of uncertainty or social organizational strategies for coping with such uncertainty. Caribbean regional organizations and states have expressed “critical concern” about “the capability to prepare fishery management and development plans” and think that “special consideration” should be given to “involving resource users in fishery management systems” (Chakalall 1992:12). This may imply that fishery planning for co-management is favoured, but it has only been assumed that the necessary social foundation among the fisherfolk exists to be built upon or can easily be induced. 2.1.3 Research questions  Using the case of the Barbados pelagic fishery, I pose the following general research question: what  potential implications do the uncertainties that fisherfolk perceive, and the  particular social strategies they adopt to cope with these uncertainties, have for fishery planning? I propose to answer this general question by addressing the three components: What uncertainties are perceived to characterize the Barbados pelagic fishery? Why are particular social strategies used to cope with these uncertainties? What potential implications do these perceived uncertainties and coping strategies have for fishery planning? This applied research cannot provide all the socioeconomic information required for fishery planning in Barbados, but it will contribute a useful, and presently absent, perspective that could be described as bio-socio-economic. In a broader context, since the type of fishery found in  9  Barbados is not often reported on, this research contributes an important additional case to the international fishery planning and management literature. Also, more information is needed on the vulnerabilities of co-management, and the conditions under which it can arise, in order to build a theory of co-management (Pinkerton 1989, 1994). This research contributes to that important effort. 2.1.4 Key concepts  This research is interdisciplinary and based on a conceptual framework, rather than a single body of theory. The concepts already introduced but not explained, which form the basis of the research questions, are explained below. Those less frequently found in the fisheries literature receive most attention, and the explanations include references to literature in which the concepts are discussed in some detail. The relationships between the concepts are addressed in the next section. Perception is the process of selecting, organizing and interpreting data to create an understanding of the environment that is meaningful to the perceiver. The process may be affected by cultural, social, educational, experiential and other variables (Downey and Slocum 1975). Uncertainty is an actor’s lack of information about, or inability to predict, a particular state or event (Cashdan 1990). This is broader than the sense used in decision theory (Connolly 1980). It accommodates the reality of the research situation. Coping is the ability to identify and pursue goals under conditions of uncertainty (Lerner 1980). Strategy is the conscious decision, involving long term perspective, to take action (Crow 1989). Social strategies include the decisions of individuals to form, or not to form, social relations with other individuals.  10 Social atomism is the assumption that society comprises independent individuals, each of whom carries out economic action to achieve goals independently arrived at, and whose aggregated actions comprise the functioning of the social system (Coleman 1990). Social embeddedness means that economic action, outcomes and institutions are affected by individuals’ relations, and by the structure of the overall interactive network of relations that integrate the individual into society (Granovetter 1985). A social network is a conception of social structure as patterns of relations (ties), with various contents, joining defined social actors (nodes) which can be individuals or collectives such as organizations (Berkowitz 1982; Marsden 1990). Individualistic social strategies emphasize the precedence of benefits to the individual over those to any collective, but may be based on either social atomism or networks (Hinde and Groebel 1991; Jentoft and Davis 1993). Organizations are social collectives predominantly oriented to the attainment of specific goals (Haralambos 1985). Social capital is a resource embodied in a network relation that facilitates an actor’s actions (e.g., obligations and expectations, aid, information) (Coleman 1990). Social power is the capability of actors, singly or collectively, to mobilize resources and carry out particular control activities in relation to other actors (e.g., orienting, sanctioning, constraining) (Burns 1985). A power resource is anything an actor may use to influence or regulate the decision-making and action of other social agents. For example, information is a potential source of power for influencing perceptions, preferences, evaluations and actions (Burns 1985). Socioeconomic indicates that social factors influence, and are an integral part of, economic objectives and activity (Charles 1988).  11 Fisheries planning is the process of making proactive decisions about alternative ways of using available resources to manage or develop a fishery with the aim of achieving goals specific to the situation. The resources are commonly human, financial, informational and physical. The goals may be biological, social, economic or political (MacKenzie 1974; Panayotou 1982). Management is the mechanism for optimally utilizing resources (Johnston 1992). Development generally implies induced positive social change, although formal definitions differ greatly (Orbach 1986). Bioeconomic means an approach to fishery planning and management structured by the state, and dominated by biological and economic theory. The state avoids power-sharing with the fishing industry which it assumes to be socially atomistic (Jentoft 1989; McGoodwin 1990). Co-management means an approach to fishery planning and management jointly structured by the state and fishing industry, and not dominated by biological and economic theory. The state engages in power-sharing with a community or organization(s) of stakeholders characterized to some degree by social networks (Pinkerton 1989, 1992). 2.1.5 Relationships between concepts  Since the essence of an interdisciplinary thesis is to integrate disciplines, concepts are drawn from several bodies of theory. This section shows how the interdisciplinary approach selected for this study is particularly well suited to the research problem. It is, in a sense, a thematic outline of the literature review in the next chapter. Since the operational use of the concepts is dealt with in detail there, below I focus on their integration into a coherent conceptual framework. The major debate in this thesis involves the applicability of bioeconomic versus co management perspectives on fishery management planning. I hesitate to label the debate “theoretical” since, as Brox (1990) argues, the bioeconomic perspective is more akin to  12 analytical modeling than to empirical explanation, whereas co-management has not yet been sufficiently researched to achieve the status of accepted theory (Pinkerton 1994). Nevertheless, these represent fundamentally important and different ways of conceptualizing relationships in the fishery as outlined earlier and, because of their differences, a first step is to identifj common ground from which the analysis can proceed. This common ground is the problem of the fisheries-related uncertainties faced by fisherfolk, fishery planners and managers (Wilson et al. 1994). These uncertainties originate from the bio-physical environment (e.g., fish biology and behaviour, climate, oceanography), the socioeconomic environment (e.g., competition, pricing, politics) and their interaction. I begin by documenting how fisherfolk perceive uncertainty, and move from there to examine their social strategies for coping with uncertainty. A person can either face uncertainty independently, or he or she can make socioeconomic connections with others who can assist. Such assistance can be used for individualistic or collective purposes, and may be formally or informally organized. Which strategy dominates at any particular historical juncture is an empirical question. A host of factors (e.g., personal income, socio-cultural history, political ideology, world events) promote or inhibit one strategy over another depending on the relative strengths of their influence. These factors independently, and in association with the choice of social strategies, affect fishery planning in terms of determining what approach is most feasible at any given time. More specifically, because state planners face a situation of natural and social scientific information deficiency, but need information on the fishery sufficient to provide initial guidance for planning, a start can be made by investigating what measurable fishery phenomena are perceived by fisherfolk to be most uncertain. This can highlight aspects of the fishery that fisherfolk and planners may find most troublesome, and to which most attention should be paid in management planning (Ruddle and Chesterfield 1976). But whether planners generally  13 perceive uncertainty as a phenomenon to be controlled or adapted to could also partly determine the management planning approach they adopt irrespective of the specific findings on uncertainty. Holling (1978) and Lamson (1984) note that fisherfolk, when allowed to devise their own methods, typically develop flexible strategies for coping with uncertainty. Hence they argue, on the basis of ecological theory, that state fishery planning and management should also be geared towards flexibility of response rather than stability through control. Wilson et a!. (1994) also use ecological theory to address this point directly. They show that experience with the often rigid, state-structured bioeconomic approach has not been encouraging, and argue that decentralized, community-based approaches to fishery management could be more flexible, and hence successful. To support their claim they draw upon evidence from maritime anthropology and sociology where the social aspects of fisheries receive attention. In this literature, however, the “co” in co-management has tended to be more about coping strategies within the resource-using mmunity rather than about cQoperative management with the state (e.g., Dyer and McGoodwin 1994). But Pinkerton (1989) points out that for co-management of highly migratory pelagics to be successful, both cooperation among resource users and cooperation between the resource users and state(s) are essential under the new legal-institutional ocean regime of the Law of the Sea. Yet, situations in which there is little cooperation among users or between them and the state, and in which state authority and management capability are also weak, have not received theoretical attention. It is possible in these cases that neither the bioeconomic nor co management approaches are immediately applicable. One needs to assess critically, on the basis of empirical data, which approach is most feasible given the social structure of the fishery, and the capability of the state. To progress, one needs to examine the actual and potential strategies of fisherfolk and the  14 state for coping with uncertainty, either separately or interactively. This requires a perspective that can address a wide range of relationships at various scales. The chosen perspective of network analysis arose in social anthropology, but its current application is multidisciplinary. It has been described as “a comprehensive paradigmatic way of  ...  studying directly how patterns  of ties allocate resources in a social system” (Weliman 1988:20). Wellman goes on to say that “its strength lies in its integrated application of theoretical concepts, ways of collecting and analyzing data” (p. 20), and consideration of social action and power (Burns 1985). Because network actors can range from individuals to states, the frill scope of fishery-related relationships can be explored. Attention can be paid to both the absence, presence and nature of relationships on the empirical basis crucial to applied research. The findings should show whether conditions favour the bioeconomic, co-management or some other approach. Since the network perspective facilitates examination of both horizontal and hierarchical relationships, attention also needs to be paid to integrating what happens at sea and ashore. The absence of adequate bio-socio-economic theory is conspicuous. Davis (1984) points out the prevalence of socioeconomic relationships within and between the harvest and postharvest sectors which influence the interaction between the harvest sector and fishery resource. The bioeconomic perspective tends to ignore such relationships, and Matthews (1993) has shown how the adoption of this narrow perspective by the state conflicts with the more holistic view of the fishery generally employed by fisherfolk. However, the co-management approach is only likely to be superior if it is specifically designed to include a broad range of issues beyond the biological and economic. This concerns theories of planning in the context of the relationship between fishery management and development planning. It includes determining which elements and stakeholders are necessary for incorporation into the fishery system for planning purposes,  15 without contributing unnecessarily to complexity and uncertainty in the planning process (Johnston 1992; Nielsen 1994). There is no universal formula that can be easily applied since responsive planning arrangements will be determined by the requirements of the specific situation and the host of factors that constitute the planning environment. My research effort aims to provide sufficient research data on uncertainty and social coping strategies to generate information that could be applied to solve the problem of determining the most feasible introductory planning approach and process in the case of Barbados, with application to other fisheries. My argument for integrating these diverse concepts from ecology, sociology, anthropology and planning is similar to that of Davis (1984) and Matthews (1993) among others. It is that theoretical concepts should be used as necessary to help us understand and solve real, complex problems. Theory should not be allowed to constrain applied research by demanding an unrealistically abstract perspective. The review of the literature in the next chapter is organized to provide additional details on the relationships between the above concepts. It will be shown how pervasive uncertainty necessitates various social strategies for coping, and how social relations and organization can be consequential for fishery planning.  16 3.  LINKS TO THE LITERATURE  By reviewing general literature, and work specifically on the Caribbean, this chapter examines the linkages between the concepts and arguments used in this study. I review first the literature underpinning the purpose of the study, then subsequent sections deal with topics arranged in the order of the research questions. 3.1 BloEcoNoMics VERSUS CO-MANAGEMENT 3.1.1 Bioeconomic model with the fish resources being common property, fishers individually are rational and try to maximize their catches, but collectively are irrational, take the [development] incentives, and deplete the stocks through overfishing” (CFRAMP 1992:10) Taken from the report of a workshop on fisheries management planning in Caribbean countries, this extract reflects a bioeconomic perspective on the fishery. No empirical evidence from the Caribbean is given as the basis for the choice of this perspective. It is presented as a self-evident, universal truth. The fishery as an open access common property resource was first addressed by economists (Gordon 1954; Scott 1955) in combination with a simple biological model (Schaefer 1954). Later, Hardin (1968) brought the common property model to the fore with his essay on the “tragedy of the commons”. Hardin misused  the term ‘common property’ by referring to  examples of communal ownership (res communes) instead of non-ownership or open access (res nullius). The debate over what constitutes common property (e.g., McCay and Acheson 1987; Berkes 1989) will not be reviewed here. Scott (1979) provides a useful review of the early development and use of the bioeconomic common property model in fisheries science. It is the basis of conventional wisdom in fisheries planning today. In essence, people in the harvest sector are cast as socially atomistic, self-interested, profit  17 maximizers, driven solely by economic goals to overexploit open access resources to the longterm detriment of everyone. The roles that social institutions and norms may play in moderating resource exploitation to levels that are sustainable are ignored (Berkes 1989). The model does not provide a clear link to the market for fish, but this is covered by the assumption of microeconomic theory that social atomism is an ideal for perfect competition. Hardin’s solution to the open access common property problem for any fishery resource is authoritative state control, while economists favour the creation of market-based, state-enforced property rights in the resource (Anderson 1986). Bioeconomic prescriptions administered in reality often include combinations of the two (Marchak et at. 1987). Considerable conflict between industry and government, the loss of community spirit, and plan failure have resulted from the imposition of fisheries plans based on the socially atomistic, state-structured bioeconomic approach. This has happened particularly where social ties have been known to be strong, and resource management socially based (Davis and Kashdan 1984; House 1988; Thiessen and Davis 1988). Yet management measures such as access limitation, fishing restrictions and creation of property rights are used both in bioeconomic prescriptions that are problematic and in socially based mechanisms that succeed (Ruddle 1988). This suggests that the principles of bioeconomic management are perhaps less at fault than the approach used to devise and implement them. The critical point is that social relations, whether they hinder or facilitate resource management, tend to be ignored in the socially atomistic model. Hence the problems or failures of state-structured prescriptions are due mostly to inattention to social realities. Others have also criticized (Ludwig et at. 1993, Wilson et at. 1994) and defended (Rosenberg et at. 1993) the non-social aspects of the bioeconomic approach, but I confine my concerns to social features. Bioeconomic modeling does not prescribe the planning process necessary to formulate and  18 execute management measures. However, from a social perspective, if the model assumes inevitable social atomism in its conception of society, then in its application it is less likely to consider the implications of either the social relations found in reality or the implications of excluding stakeholders from management decision-making. The latter creates serious problems due to the resultant lack of legitimacy of management measures in the eyes of the resource users (Jentoft 1989; Dyer and McGoodwin 1994). However, if the bioeconomic approach is inappropriate, what are the alternatives? The fisheries literature provides no simple answer. There is no bio-socio-economic fishery model, and no theoretically based maximum social yield to parallel the maximum economic yield (IVIEY) and maximum sustainable yield (MSY) of the bioeconomic model (Poggie 1992). The concept of optimal yield (OY) used in the United States to include social issues has not been particularly successthl (Dyer 1994). 3.1.2  Co-management Moving away from the socially atomistic and authoritarian, state-structured constraints of  the bioeconomic approach requires a conceptual shift towards co-management, a participative planning and managing process that is not by itself a model of the social, biological and economic aspects of the fishery. Co-management is “power-sharing in the exercise of resource management between a government and a community or organization of stakeholders” (Pinkerton 1992:331). Since it incorporates, but is not exclusively concerned with, biological and economic factors, it allows the implications of personal networks and formal organizations to be considered. According to its proponents, co-management theoretically offers benefits of “more appropriate, more efficient, and more equitable management” than the conventional bioeconomic approach (Pinkerton 1989:5). It can also be argued that successful co-management comes closer to Hardin’ s recommendation of “mutual coercion, mutually agreed upon by the  19 majority of the people affected” (1968:1247) than either unilateral, state-structured regulatory regimes or quasi-property rights (Jentoft 1989). Co-management is seen as a possible route towards community-based development, state decentralization and participatory democracy especially where, among other things, resource and resource user boundaries can be easily defined (Pinkerton 1989; Dyer and McGoodwin 1994). But the conditions under which co-management is feasible have not been exhaustively researched (Pinkerton 1989, 1994), and tests across a range of social conditions are essential. One critical test is to determine if and how co-management may be achieved in a highly competitive fishery. To conduct such a test one needs to find a case or cases in which co management is being considered, but where community or collective action is not accentuated. In developing countries the state often encourages the formation of community or broaderbased fishing industry organizations through which it can channel economic development efforts, and for this narrow purpose it has sometimes been successful (Bay of Bengal Programme 1990). When such a state initially subscribes to co-management it may assume without justification that social relations in the fishery meet, or have the potential to meet, the requirements for fuller participation. Yet actual social networks and organizations, perhaps because of prior historical or development experience, may not fit this requirement (Polinac 1988b; Jentoft 1989). This underscores the need for socioeconomic information to evaluate the actual social organization and appropriate planning approach (Cycon 1986; Orbach 1986; Dyer 1994). One of the challenges of this study is to determine whether the social scenario in Barbados, which superficially seems hostile to co-management, is actually so. Figure 3.1 provides a schematic overview of my overall research framework from a social action perspective. What it shows sequentially is that the social actors in the fishing industry perceive and then try to cope with uncertainty through social strategies that have various  20 outcomes. These outcomes provide feedback, sometimes by altering the source of uncertainty itself, or the perceptions of uncertainty. Fishery planners, within the parameters of the political and bureaucratic framework, have the potential to influence all stages of this process by information or action. Conversely, what is perceived or is happening may influence planners if they have information on it. All of this must operate within the broader contexts of the Barbadian culture and society, and the international environment. 3.2 FISHERIES-RELATED UNCERTAINTY 3.2.1  Perceptions and planning  Planning involves proactive decision-making based on forecasts of events or actions. Since perceptions influence people not to act, or to act in various ways, knowing what people perceive assists in reducing the uncertainty of planning, assuming perceptions remain stable. Such an assumption is unrealistic, however, since perception is more than a deterministic stimulus response function (Downey and Slocum 1975), and change in perception also needs careful consideration.  21 Figure 3.1 Social research framework  INTERNATIONAL ENVIRONMENT  Research on perception shows how people interpret uncertainty in various ways, and how these interpretations may vary according to such attributes as age, gender, education or occupation (Institute for Social and Economic Research 1982). Perceptions that depart considerably from some accepted standard or that are lacking may indicate the need for providing education and information. If fisherfolk make frequent observations of measurable phenomena in the environment, knowing their perceptions can also assist planners in developing  22 useful estimates of changes in natural resources (Ruddle and Chesterfield 1976). How perceptions of uncertainty can be used to supplement scientific measurement is a concern of this study. Knowing the ways in which perceptions arise, are maintained and change is particularly important if perceptions are used to supplement environmental data collection. But such research would involve investigation of both cognitive and social processes (Downey and Slocum 1975), and is well beyond the scope of my research. While my focus is on  fisherfolk’ s perceptions of potentially measurable uncertainty, I  acknowledge that every part of this study concerns perceptions in some way. For example, how planners perceive fisherfolk influences whether they adopt top-down or people-centred planning. How clients perceive planners and their plans influences the effectiveness of planning and management. And how both planners and their clients perceive social science influences whether it becomes integrated into the planning process. Moreover, the absence among stakeholders of shared perceptions of the environment, of problems and solutions, and of each other poses a serious obstacle to participatory planning (Ruddle and Rondinelli 1983). Some of these factors are dealt with in this thesis, but less explicitly than perceptions of uncertainty. 3.2.2  Uncertainty is ubiquitous Uncertainty permeates the fishing industry, but understanding how people respond to it is  still in its infancy (Cashdan 1990). The fisheries literature is mostly concerned about uncertainty from the viewpoint of the fishery scientist or manager, and the way in which the uncertainty of the ecological and economic environment complicates management, especially in multispecies, multifleet situations (Hilborn and Walters 1992). My interest differs. It is primarily in how uncertainty affects fishers, boat owners, fish vendors and fish processors in their everyday activities. Uncertainty has three main interrelated sources: the natural resource environment, human  23 activity and its influence on this environment, and the limitations to human understanding of the total environment. This last point implies that in many situations we cannot distinguish between stochastic systems and complex deterministic systems that produce chaos. Chaos is uncertainty between reasonably stable bounds, but often we cannot discover these bounds (Wilson et al. 1994). Temporal and spatial scales are also relevant to the measurement and perception of uncertainty, but uncertainty may either increase or decrease with time or space depending on the phenomenon and how it is measured or perceived (May 1983). One strategy to cope with uncertainty, typical of scientists and managers, is to try to control the decision setting so as to minimize surprises (May 1983). Another strategy, typical of most resource users, is to focus on being sufficiently flexible to adapt to whatever surprises arise (Ruddle and Rondinelli 1983). The latter is of greater concern in this study as regards strategies for coping with uncertainty. Sources of uncertainty in the harvest sector, many of which are shared by both fishers and boat owners, include environment (sea conditions, oceanographic features), catch rates (per trip, per season), equipment performance (mechanical failure, gear loss), human factors (experience and skills of captain and crew, competition and conflict between fisherfolk), price (sensitivity to volumes landed, demand, exchange rates) and institutions (government regulations, labour relations) (Gates 1984). Boat owners face the additional uncertainties surrounding their financial decisions such as the borrowing of investment funds and determining the appropriate level of investment (Bort 1987). The major uncertainties in marketing seafood are associated with price and quantity of fish, periods of delay between first-hand and final sales, product image and demand changes, product quality, and interference of market signals particularly by government intervention (Prochaska 1984). Of the above, supply and price uncertainties are shared by fishers and boat owners. Through increased regulation, fishery management itself may become a principal source of  24 uncertainty. Decision-making uncertainty in the management or planning apparatus is added to the uncertainty of nature—the rules governing the fishery keep changing (Hannesson 1984; Anderson 1984). Some argue that, because of uncertainty, fisheries science cannot provide solutions to resource sustainability problems without integrating social and political aspects (Ludwig et at. 1993). Others state that while fisheries science has provided solutions under conditions of uncertainty, it has been the inability of managers to implement scientific advice that has lead to the failure of bioeconomic plans (Rosenberg et at. 1993). 3.2.3  Uncertainty in the Caribbean context In the eastern Caribbean, since the pelagic fisheries are open access and unmanaged, fishery  management plans are not sources of uncertainty for resource users. However, scientists, planners and managers (often all the same person), plus resource users, face significant challenges from the marine environment and limitations to comprehension. Perceptions of the same phenomenon may differ—fishers for example seeing seasonality in terms other than competing scientific hypotheses about stock structure and migration, but since planning involves communication, it is useful to look at the information which fisherfolk and the state may have to exchange. The value of this information as perceived by the other may determine the nature of state/industry interaction. Only a few illustrations are sketched below from the perspective of the planner/manager/scientist. Figure 3.2 shows a few of the features to be discussed.  25 Figure 3.2 Ocean environment around Barbados  Angudla  - -  Antgua/Barbuda  . .---  -  vi: Saba St. Kit:ts/ \ CV  -  S  /  GoadCIoupQ  C  )omiriIca  1artlniqU  St.  Luc i.  .;  GC  NEC GC —*  27°C  North Equatorial Current Guiana Current Current direction Mean annual sea surfacc temperature Hypothetical equidistant EEZs  /  26 3.2.3.1  Marine environment  Both fishers and planners typically need to reduce their uncertainty about the relationships between oceanographic and ecological processes, but scientists know little about the interannual variability of climate and oceanography in the eastern Caribbean. For example, although the mean monthly patterns of circulation suggest smooth current flows, recent research shows that there are complex systems of gyres and eddies both east and west of the Lesser Antilles, and uncharted upwelling may occur near the islands (FAO 1993; Bowman et at. 1993). The spatial variability of currents, eddies and water types is so great in the region that coarse sampling from ships, or short time series of observations, could lead to misconceptions about oceanographic processes (Muller-Karger 1990). There is also uncertainty about processes at all trophic levels. For example, there are conflicting reports on the causes of seasonality in zooplankton abundance (Muller-Karger 1990). While there is debate about which, if any, approaches to modeling multispecies fisheries are appropriate, understanding the dynamics of primary and secondary production is basic to most (Hilborn and Walters 1992). Such understanding is crucial to the ecosystem modeling which Wilson et at. (1994) propose as a way of reducing uncertainty, and distinguishing between complexity and chaos, by using the ecosystem knowledge of fishers to make scientific data analysis more robust. Reducing the oceanographic uncertainty described above may be possible by involving the harvest sector in sampling the marine environment at finer spatial and temporal scales than presently feasible through conventional research (May 1983). 3.2.3.2  Harvest and post-harvest  Regarding commercial fishing, the information base for the assessment and management of pelagic fisheries in the eastern Caribbean is patchy and scant. For example, migratory pelagics (flyingfish, dolphin, wahoo, kingfish, tuna, billfish, swordfish and shark) show marked seasonal  27 availability but it is currently impossible to distinguish between large scale directional migration and smaller-scale onshore-offshore movements (FAO 1993). Stock definition and the status of the stocks are also poorly known, with few scientifically rigorous conclusions (ICCAT 1992). Regarding stock assessment, the practice of different boat types targeting several species on the same trip by using a mix of gear confounds the fine tuning of effort indices (Mahon et at. 1990). Recording species in groups, gaps in the data, and unquantifiable errors confound the commercial catch time series presently available (Mahon and Rosenberg 1988). Without better information on stock structure and catches only crude estimates of potential yield have been possible using data from other, roughly comparable, tropical regions (Marcille and Caddy 1987; Singh-Renton and Nielsen 1994). Much uncertainty surrounds these estimates due to differences in the productivity and fishing effort in these surrogate areas. Uncertainty about the marine environment highlights the need for studies on how much fishers know about where and when various species are available, whether there are indicators which fishers use to determine availability, and what changes in fishing strategy are employed in response to perceived or anticipated changes in availability (Mahon et at. 1990). The sharing of perceptions and measurements of harvest sector uncertainty seems to be a necessity rather than a luxury in this predicament of information paucity. Although uncertainty affects not only the harvest sector, very little has been written on seafood marketing and distribution in the eastern Caribbean except for acknowledging that it is poorly developed and a major constraint to continued economic growth and fish consumption (Systems 1984). The uncertainties that affect the fisherfolk (all participants in the fishery) are unknown, and as with other small firms, lack of information on the productivity and profitability of fishery operations has made it difficult for state planners to evaluate and address the perceived problems (Howard 1989).  28  3.2.3.3 Geo-politics While the eastern Caribbean states have not yet engaged in management planning for their pelagic fisheries, they face several geopolitical uncertainties which in turn affect their fishing industries. First, the pelagics are shared within the region and with high seas international fisheries. Evidence for the assumption that stocks of some species probably migrate only within the Lesser Antilles sub-region, is largely circumstantial (FAO 1993). Furthermore, there are few negotiated marine boundaries delimiting exclusive economic zones (EEZs). If EEZs were to be established, the region would be carved into a complex patchwork with none of the national extended jurisdictions matching appropriate management units (Mahon 1 990a). This causes uncertainty about where boats may legally fish, and how regional and foreign fishing will be managed, under the United Nations Convention on the Law of the Sea (UNCLOS). With some of these issues being addressed mainly by state and research agencies, the fisherfolk may be unaware of the political and scientific complexity. Yet, as stated at the outset, their perceptions of the situation are important especially if fisherfolk wish to adapt to or control aspects of uncertainty by playing an active role in planning. Uncertainty generated by man-made complexity may not be tolerated in the same way as that from the natural environment which, according to Aronoff (1970), Caribbean fishers enjoy as a valued feature of their work. 3.2.3.4 Summary  In summary, uncertainty is as pervasive in the pelagic fisheries of the eastern Caribbean as elsewhere, and the uncertainty is mainly in the form of widespread absence of information. The task is first to identify which uncertainties are perceived in Barbados to be of sufficient importance that strategic social responses for coping are warranted. It is also necessary to determine how perceptions of uncertainty differ among people in different positions in the fishing industry, and differ from measurements of the phenomena perceived to be uncertain.  29  How fisherfolk cope with uncertainty through social mechanisms is examined next. 3.3 SOCIAL STRATEGIES FOR COPING WITH UNCERTAINTY 3.3.1  Social atomism versus embeddedness Bioeconomic models assume, among other things, that economic behaviour is socially  atomistic such that social relations are either irrelevant or non-existent (Figure 3.3a). This undersocialized conception of economic behaviour is crucial to the paradigm since, according to Adam Smith, social atomism is a prerequisite for perfect competition. To conceive of economic action as social action (i.e., as relational) weakens certain rational choice assumptions such as self-interest, individual utility maximization, and the rationality of organizations. Particularly because of deficiencies in explaining empirical evidence, rational choice models have been criticized, and alternatives have been presented (Zey 1992). The theme of social embeddedness is an alternative to the atomism assumption for explaining economic behaviour. According to the embeddedness argument, economic “behaviour and institutions are so constrained by ongoing social relations that to construe them as  independent is  a  grievous  misunderstanding”  (Granovetter  1985:482).  However,  embeddedness is not equivalent to the functionalist over-socialized concept of human action, where individuals are assumed to be governed or pre-programmed entirely by norms or customs alone. By removing any active role for social relations this over-socialized conception reverts, in effect, to atomism (Granovetter 1985). Compared to atomism, many argue that embeddedness appears to better explain empirical evidence from sociology, anthropology and economics (Wellman and Berkowitz 1988; Plattner 1989; Zukin and DiMaggio 1990). This is my rationale for using social embeddedness as a key concept in this research. In turn I use the term “socioeconomic” to emphasize my perspective of economic activity being inextricably embedded in its social context. Discovering whether  ,.,  .3  working from the perspective of embeddedness is consequential and useful for planning and managing a fishery, or whether, in contrast, social atomism assumptions would suffice and be more useful, is at the heart of this applied research. As discussed above, atomism corresponds closely to the western concepts of individualism or autonomy involving the shedding of social attachments. In contrast, the Afro-Caribbean concept of autonomy is of an individual realizing independence through strong interpersonal support (Sutton and Makiesky-Barrow 1981). Thus individualism or autonomy conveys an impression of networks rather than atomism in the Caribbean. However, since individualism often implies that each person tends to act in their own interest, it may be functionally similar to atomism for the purpose of planning and management. Social networks and organizations  3.3.2  The social network concept operationalizes embeddedness and is the foundation of network and structural analysis (Berkowitz 1982). My interest is in the personal or egocentric networks of people in the pelagic fishing industry where the fishery actor, as the focal individual (centre, ego) of his or her personal network, uses the social capital embodied in ties to other actors (alters) in strategies for coping with uncertainty or is constrained by ties to other actors (Figure 3.3 b).  3  Figure 3.3 Social atomism, networks and formal organization  (a) Social atomism  (b) Social network  BO= boat owner CA= fisher captain KEY CR= fisher crew FV= fish vendor FP fish processor = ties between alters and egos; and = links between alters provide social capital = extent of formal organization within which members are engaged in collective action, enabled and constrained by their ties and links to others within and outside of the oranization  (c) Formal organization  Not all ties are beneficial to the focal individual, and both constraining and enabling ties are of interest in this study. I am also interested in the shared attributes of individuals associated with characteristic patterns of relations, and in how social relations differ. My approach is related to research on the use and exchange of material aid, advice, financial assistance and the  32 like in personal networks for social support (Weilman et at. 1988; Weilman and Wortley 1990), but I employ network analysis in a limited way. I look only at specific enacted informational and instrumental ties, and I also use attribute, not relational, categories to classify people (survey respondents) in the industry. Network analysts criticize the methodological individualism typical of social surveys where respondents’ categories, and their expected roles in social organization, are determined solely on the basis of their attributes, not their relations with others (Weilman 1988). However, in practice, fishery planners depend heavily on knowing the categorical memberships of people in the fishing industry  —  in distinguishing fishers from non-fishing boat owners, for example. Yet,  while it is necessary to categorize the industry by such occupational labels, it is also useful to acknowledge the network argument that persons with different categorical memberships may perform the same functions in a network. For example, both fishers and non-fishing boat owners may supply the same information on fish catches to a focal individual, making their categorical memberships immaterial for this purpose. This suggests that planners could benefit from having information on both categorical and network memberships, and that the similarities and differences between the two would be of practical interest. Social networks are said to be universal (Bott 1971). But Wilson (1973), in doing research in the Caribbean, warns that when the network is used as an analytical research construct it should not also be assumed to constitute social reality for the people studied. That is, they may not consciously perceive networks (at least not in the manner of the researcher) or operate purposefully on a network basis. For them society may be organized in a different manner, causing them to pursue a different strategy for coping. The formal organization is a social unit that, from the embeddedness perspective, is superimposed upon the personal networks of the individuals that are its members (Granovetter  33  1985) as shown by the shaded area in Figure 3.3c. I use ‘formal’ to imply an organization with a name, elected officers and perhaps legal identity such as a union or co-operative, thereby distinguishing it from a network. An organization can also be a node in a network with other collectives, and in which social relations and power constrain and enable the activities of organizations just as they do individuals. Similar to the refutation of social atomism, this view contradicts the rational choice perspective that organizations function rationally, but at the same time it appears to more consistently explain empirical findings (Zey 1992). Sampling Caribbean sociology  3.3.3  Much of Caribbean sociology focuses on social stratification, with M.G. Smith’s image of a plural society currently being most influential (Craig 1982). It is not my intent to provide an exhaustive accounting of how culture, race, colour and class may have shaped Caribbean societies (Smith 1984), but I necessarily rely on some historical detail to set in context certain explanatory accounts of the current situation. In this section I draw selectively upon the small body of literature dealing with social networks in the Caribbean. Most of this literature is from writers concerned with female centered kin and non-kin supportive networks in the lower or working class. Powell (1982), in reviewing the use of the network concept, notes that much of the work showing West Indian women in social networks does not use the network approach and terminology. None of this literature is specific to the fishing industry, and male networks are mentioned only occasionally. Kinship, individualism and gender are variables commonly discussed. 3.3.3.1  Kinship  Rodman (1971), researching the black lower class in Trinidad, found the concept of kinship to be open with indistinct breaks and exceptional fluidity between nuclear family, household, extended kinship and community relationships. The kinship group did not engage in corporate  activity, and kinship ties were used primarily for security and survival. The use of fictive kin labels for non-kin often signifies the strength of a relation (Powell 1982). More recently, Le Franc (1989) examining higglering (small-scale trading) in Jamaica, also found that the family enterprise, family based economic networks and family labour dependence were generally absent. Recruitment to the enterprise had little to do with family ties and traditions, and family or household members competed in the same economic activity. One cannot readily use the family or household as an analytical unit due to frequent changes in household composition and individualistic relations among household members (Rodman 1971). That is, in general, the personal aspect of relations is more important than kinship. A finding of weak kinship ties among Caribbean fisherfolk would be contrary to most social science accounts of small-scale fishing industries (Acheson 1981). The favoured hypothesis is that the weakness of formal kinship ties and importance of the personal element give the individual greater flexibility in adapting behaviour to circumstance (Rodman 1971; Barrow 1986). In the fluctuating economies of the Caribbean the most successful strategies are said to be the most flexible ones. So a flexible network of kin and friends is an adaptive social feature (Gussler 1980). By selectively opening, reinforcing and terminating dyadic ties with kin and non-kin, an individual can maximize opportunities to conserve or increase scarce resources, and minimize social obligations or expectations that deplete resources (Gussler 1980). Individualism has consequences for the form of the network and how individuals manage their networks. 3.3.3.2  Individualism  Mintz (1971) observed that the main features of social interaction among rural lower class people are dyadic ties formed along lines of common interest to satisfy particular individual needs. This implies either a predominance of dyadic ties, not true networks, or networks with few or weak links between members. Consequently, individuated social structure may require a  3  person to have a large number of dyadic ties in order to receive adequate information and support. A relationship with one person will not naturally lead to ties with their circle of friends and relatives as is probable with well developed personal networks (Gussler 1980). Gussler found in St. Kitts that if you supplied a person with a valued resource, in order to monopolize it the person would discourage you from having additional relations with friends and foes alike. However there is insufficient evidence to say whether this non-sharing behaviour and the predominance of dyadic rather than network ties can be generalized to all countries and sections of society in the Caribbean. Other authors have sought to explain Caribbean individualism not only by present circumstance, but also by historical political economy. Mintz (1971) argues that the plantation served as a societal model which individualized Caribbean peoples, having greater effect in places such as Barbados where peasants competed for scarce non-plantation land after slavery. According to Makiesky-Barrow (1976), plantation organization constrained collective action and encouraged individual efforts at improving life chances, but individual advancement on the plantation implied accommodation to planter interests in opposition to slave interests. Unity or solidarity was experienced mainly in collective, egalitarian resistance to subordination and was viewed as a strategy for change. She argues that this provides the historical basis for distrust of higher status groups and the perception that status differences threaten group interests. It implies that individualism is more than a response to current circumstance, but is culturally rooted in history and hence enduring. However, several authors have noted in Caribbean societies a dynamic duality or tension between a western liberal type of ideology, espousing individual achievement to seek higher status, and a more socialist ideology stressing unity and equality (Rodman 1971; Wilson 1973; Makiesky-Barrow 1976). This duality affects individuals, networks and organizations. Among  islanders whose networks were based on egalitarianism, Wilson (1973) describes how a local higher-status change agent was ineffective at organizing and effecting change because he did not subscribe to their system of values, and hence lacked social acceptance and authority. This seems relevant to fishery planners’ interactions with fisherfolk since the two are likely to be of different social status. More important, if fisherfolk are extremely individualistic, then social atomism may come closest to modeling frmnctional social reality, and collective cooperation may be unrealistic. 3.3.3.3  Gender  Finally, networks seem to differ by gender. While lower class women in the Caribbean perceive themselves as having a hard life, they also live in societies where their self-reliance and competence go relatively unchallenged, and there is less gender inequality than in western industrialized countries (Bourguignon 1980; Sutton and Makiesky-Barrow 1981). This too is a product of the plantation past where there was little sexual division of labour and the adoption of colonial gender differentiated culture was discouraged (Hart 1989). If there is generally less inequality, gender differences are of particular interest. A high percentage of West Indian households are female-headed, and women are more active in maintaining extended kinship and non-kinship networks. These they use as sources of assistance and support for activities such as child-rearing (Powell 1982). Rotating credit associations (vestiges of West African society) common in the lower class throughout the Caribbean as informal network-based mechanisms for saving usually have a higher participation rate among women than men (Barrow 1976). Also, women’s networks are said to be more dispersed geographically and less homogenous (in terms of members’ social status) than men’s networks (Sutton and Makiesky-Barrow 1981), and while reciprocity is strong in both, it is more balanced for males than females (Powell 1982). Small all-male groups are common in  37 Barbados (Stoffle 1969) and elsewhere in the Caribbean (Wilson 1973). They tend to be egalitarian, more social than instrumental, and have boundaries that are fluid, with membership changing frequently (Makiesky-B arrow 1976). 3.3.3.4  Summary  In summary, the use of social networks is a survival strategy of the Caribbean black lower class, but such networks exhibit gender differences and are based more on individualistic personal relations than kinship. So although the network concept has not been applied to eastern Caribbean fishing industries, with most fisherfolk being from the black lower class, the above findings should be relevant for comparison. While, as in the research reviewed above, the network concept tends to be used more as a metaphor than as an analytical approach, the nonCaribbean fisheries literature provides useful examples of the ways in which networks function in the fishery. Some of these are reviewed below. 3.3.4 Social strategies in the fisheries literature  Evidence from reviews of the anthropology and sociology of fishing shows that social ties are used to cope with uncertainty in the fishing industry (Acheson 1981; Pollnac 1988a; Bailey et at. 1986). The literature on the eastern Caribbean does not deal with perceptions of uncertainty or social strategies for coping, yet the ubiquity of these phenomena leads me to suspect that they may be found there. Given the thorough reviews mentioned above, only a brief synopsis is provided here referring to Figure 3. 3b as a generalized example. 3.3.4.1  Networks  In this example, with the boat owner as the focal individual, recruiting fishers on the basis of kinship obligations secures a core supply of available and hopefully trustworthy labour (Thiessen and Davis 1988). If additional fishers are contracted, their labour may be tied as expected repayment of credit advanced by the boat owner during periods of low catches (Platteau 1984).  .3  The fishers, in turn, form networks to reduce the uncertainty at sea by exchanging information on sea state, fishing success, etc. between members and concealing it from outsiders (Palmer 1990). Information is more likely to be shared in a fishery for migratory species (Stuster 1978). In my illustration the boat owner provides the fish vendor and processor with supplies of fish on a regular basis in preference to other buyers. In return she receives fishing inputs from the processor (Barrett and Apostle 1989). Both she and her fishermen receive small loans from the fish vendor which may be especially valuable when maintaining or modernizing the vessel (Bort 1987). However, credit ties can be exploitative, conflictual and perpetuated by buyers (Bailey 1983). Yet fishermen and fish buyers who do not form ties are at a competitive disadvantage (Wilson 1980). Not shown here are the collusive ties for price-fixing and non-market trade between buyers which reduce their uncertainties about market fluctuations (Peterson and Georgianna 1988). Hierarchical ties such as those found between the layers of the harvest and marketing sectors have received special attention since their prevalence is not explained well by neoclassical economic theory (e.g., Williamson 1975), but become clearer under the social embeddedness theme (Granovetter 1985). Of particular interest to this study is the relationship between social ties and fisheries management planning. No generalization is possible. Although social ties, often on a community basis, have served fisheries management functions either intentionally or as byproducts of traditional custom, they also have been used exclusively for increasing the efficiency of exploitation (McCay and Acheson 1987). In some cases socially based resource management on its own has been very successful (Johannes 1978). In others it has co-existed peacefully with official regulations (Berkes 1985). In an instance where social ties were used to facilitate illegal exploitation, self-regulation was rejected in favour of imposed external policing (Taylor 1987).  -3  Most often, however, where social ties have been evident, management failure, conflict between industry and government, and the loss of community spirit has resulted from the state’s imposition of fisheries management plans based on the socially atomistic model (Davis and Kashdan 1984; House 1988; Thiessen and Davis 1988). According to Jentoft and Davis (1993 :3 60), modernization and use of this model in eastern Canada has generated among small-  scale fishers an ethic of “utilitarian individualism” typified by “profoundly self-interested” goals and strategies, “primarily economic in character”. 3.3.4.2  Organizations  In developing countries, cooperatives are the most common form of fisher organization (Pollnac 1988b; BOBP 1990). Much less has been written about formal collectives in the postharvest sector, presumably because these are less common for two main reasons. First, since fish buyers are, or are assumed to be, better off than fishers, they seldom receive the same level of state attention in terms of organizational promotion as an aid channel. Also, as described by Emerson (1980) and illustrated in Figure 3.3c, vendors and processors may be able individually to gain sufficient control over the harvest sector, even those members in organizations, through network ties that penetrate formal bodies, that their own collective is unnecessary (Meynell 1984). The greatest barriers to fisher cooperative formation and successful operation are often said to be such penetrating ties especially where the state introduces the cooperative idea as an alternative to dealing with exploitative middlemen. In this situation the fisher who joins the co op is branded, not as an individual exercising alternatives in an open market, but as a traitor to customary relations and community structure (Emerson 1980). Another barrier is individualism, with utilitarian individualists who evaluate fisherfolk organizations mainly in instrumental, profit oriented terms being unlikely to exhibit a durable cooperative ethic (Jentoft and Davis 1993).  40 Jentoft (1985) sees the fisherfolk co-ops as underutilized tools in fisheries management planning. Yet, fisherfolk organizations in developed countries are also prone to conflicts and conspiracies, especially where the body is put in a position of power by the state as in co management (Jentoft and Mikalsen 1987). The strong ties formed between the elite in the organization and state officials, compared to the weak ties between the same elite and lower ranking members are of particular concern in a case such as the above due to the enhanced possibility of subtle or overt state manipulation. This scenario is raised again when reviewing some of the planning literature. 3.4 PERSPECTIVES ON PLANNING  In this section I take a comprehensive look at planning in general, and in the Caribbean context, in order to situate this study within the theoretical literature and practical application on the basis of experience. 3,4.1 Information as a social power resource  The social power of organizations and persons occupying positions within them can be derived from individual action through the network concept. This is essentially the case when citizens elect a government which then delegates some aspects of the exercise of power to its agencies and officials. In this manner persons in the fishery planning apparatus acquire power over aspects of the fishing industry. They may use information about the fishing industry on its own to change perceptions, and with other resources as power for sanctioning or constraining actors (Burns 1985). Yet conventional biological and economic fisheries studies ignore the potential use of their information output as a power resource since the concept of power, in the political context, is not considered in either discipline. But applied scientific information is neither neutral nor value free. For example, while major reductions in fishers’ perceptions of uncertainty usually originate  41 from the experience of fishing itself, fishery planners and managers can assist this reduction through the provision of information (e.g., biological surveys, correlation between fish abundance and oceanographic features, sea state forecasting) if they are aware of perceived uncertainties (Mangel and Plant 1984). Conversely, withholding information or allowing false information to persist can exacerbate uncertainty and consequently perhaps slow fleet capitalization or other activities. Differential access to information can result in power differences within the industry, and fisherfolk may use their information to bargain with the state (Ruddle 1994). Whatever is the case, the socioeconomic implications of disclosing or revealing information should be known to planners in order to inform decision-making (Peterson and Smith 1982). Action arising from such decision-making involves the exercise of power by the state in the context of the overall social setting depicted in Figure 3.1. My concern is with the potential exercise of power since the actual extent of power is only revealed upon application. That is, the activities in which actors participate contribute to maintenance or change in the institutional framework and relative powers among actors (Burns 1985). An investigation of revealed power dynamics is beyond the scope of this study since the  state did not undertake fishery planning during the period of fieldwork. 3.4.2  General considerations The finding that failure in fishery planning has not been in all cases a fishery problem, but  sometimes a reflection of problems with planning in general, has been reviewed by Rothschild (1973) and MacKenzie (1974). While not suggesting that interdisciplinary research and problem solving is the only answer to these problems, they suggest that a narrow perspective constrains appropriate decision-making. Their other comments will not be repeated except to note that failure to understand the linkages in the total planning environment, failure to involve clients in a continuous iterative process of plan review and refinement, and failure to take into account  42 complexity and uncertainty figure high on their list of identified problems. A substantial part of planning involves decision-making under uncertainty and devising strategies for responding to uncertainty by adaptation or control (Conyers and Hills 1984). Decision-making under the uncertainties facing scientists, managers and planners has figured prominently in fisheries literature and in the recent debate on the role of scientists and scientific advice in planning for fishery resource sustainability (Ludwig et a!. 1993; Rosenberg et a!. 1993). However, although uncertainty as a statistical problem for planners and managers must be appreciated, this aspect is not as important in this study as its social consequences for fisherfolk and state officials. While I have usually referred to fishery planning as if it were a single topic, it comprises both management and development planning, and the relationship between the two components needs to be examined. According to Chakalall (1992:56): “a fishery development plan should not be confused with a fisheries management plan, which focuses mainly on resource conservation and forms part of the development plan.” Johnston (1992) locates fishery management as a development strategy and proposes that in unmanaged, open access fisheries, fishery development should take the form of fishery management. This is to allow decisionmakers to acquire a “thorough knowledge of the resource, economic conditions in the fishery and elsewhere, and the legal/institutional environment” prior to embarking on development planning (Johnston 1992:11). However, the political reality is that fishery management usually follows development, and is supported mainly in order to improve the lot of stakeholders already engaged in development who are experiencing problems (Caddy 1984). Thus the relation between management and development planning may be more complex in practice than in theory and must be taken into account in applied research. Although the same planning process can be applied to both fishery development and  43 management planning, I focus more on the latter. This is for two reasons. First, the priority in the eastern Caribbean and elsewhere is for examining the options for managing fisheries (Mahon 1990a). Second, this focus is consistent with the notion that Caribbean planners, and public servants in general, should cultivate a sense of managing rather than administering aspects of development (Demas 1992). 3.4.3 Integrating social science into fishery planning  The argument for integrating social science, specifically sociology and anthropology, into fishery management planning rests on the fact that, whether or not they affect fish, the implementation of fishery plans inevitably involves social change. Orbach (1986:105) uses the cliché that “one does not manage fish, one manages people” to point out that social information may be useffil at all stages in the planning process. This argument is strengthened by what authors have been calling a “crisis” in world fisheries management (McGoodwin 1990: Nielsen 1994). The situation is that resource overexploitation has not been halted by bioeconomic restrictive control or market-based measures partly because in the eyes of resource users these regimes lack legitimacy. The solution tendered is that legitimacy and overall effectiveness can be improved by greater involvement of resource users in management, transferring some management responsibility to them if possible. This is the growing argument for co management. If disenchantment with bioeconomics promotes social considerations and co-management, then it is pertinent to ask first if social science can be successfully integrated into fisheries management planning and then what useful perspective can it provide. I attempt to answer these queries by examining some of the literature written from a social perspective on participatory planning in developing countries and on co-management. Throughout I try to show the relevance of the concepts introduced earlier, especially the network concept.  44 34.3.1  General problems and politics  Problems encountered with the integration of social science often relate to the history of fishery management planning. First, since sociology did not accompany biology and economics in the evolution of fishery science, social accounts of fisheries are seldom in the form of applied research immediately usable by planners, even though the number of such accounts has been increasing (Dyer and McGoodwin 1994). Second, with few persons inside fishery planning agencies having a background in sociology or anthropology, it is difficult for such agencies to collect social data and convert it into useful information (Ward and Weeks 1994). Third, being accustomed to quantitative bioeconomic methodologies and prescriptions, fishery planners find it difficult to accept and integrate the more qualitative aspects of social science (Orbach 1986). Especially in multispecies fisheries, experience has shown that social issues fail to be included due to the existence of already complex biological and economic interactions (Dyer 1994). Such complex fishery interactions are prevalent in the Caribbean. In terms of the mandatory incorporation of social issues into fishery planning, the United States’ Magnuson Fishery Conservation and Management Act (IVIFCMA) of 1976, with its concept of optimum yield mentioned earlier, is a frequently examined model. However, although provision is made for social impact assessments, appointing social scientists to advisory panels, and using a system of public hearings to broaden participation, social and cultural factors are still not being integrated into management decision-making (Smith 1982; Paredes 1985). Dyer (1994) elaborates on the barriers to achieving effective fisheries management planning under the U.S. system and suggests some strategies for overcoming these. A few of the strategies most relevant to this study include more balanced representation from user groups, better communication between stakeholders, the co-management of resources, and more proactive planning involving social scientists. At present fishery planing is reactive, “the need for  45 regulation is generated by perceptions of users and policy makers  ...  instigated by declining catch  levels, economic factors, politics or other considerations” (Dyer 1994:84). He admits that politics often overrides science in the final stages of planning. The inclusion of politics should not be surprising since planning is fundamentally a political process. Plans are political statements, and technically superior plans that are politically unacceptable usually gather dust. Political theory provides a basis for the legitimacy and form of planning (Low 1991), and is relevant to this study since social embeddedness also implies that relations are influenced by politics (Zukin and DiMaggio 1990). Western liberal-democratic politics, and atomistic, market oriented economic ideology are often associated. Yet liberal politics and market-based economics may not necessarily rule out alternatives to the bioeconomic approach, the most obvious being co-management. Where the maxim “less government is good government” is put into practice, it is conceivable that the state may wish to delegate responsibility and encourage co-management of the fishery (Jentoft 1989). 3.4.3.2  Summary  In summary, before one can examine the implications of social information for fishery planning, one must consider how that information is going to enter the planning process. It appears that there have been difficulties with the integration of social science information even when institutionalized through legislation. However, liberal-democracy as an ideology does not appear to be a factor inherently in opposition to social science integration or co-management. Pinkerton (1989, 1994) and others discuss several additional factors that influence the establishment of co-management regimes. 3.4.4 Co-management and participatory planning  In this research I focus on the influential factors most relevant to the perspective that formal organizations comprise social networks which link members to each other and to the outside.  46 Most significantly, the institution of co-management typically requires that fisherfolk organizations participate with the state in the management process. This necessitates considering the concept of participation in the context of citizen/state interaction. There is a large body of literature on participation, and the concept has been reviewed generally (Oakley and Marsden 1984) and in the context of fisherfolk (BOBP 1990) especially in relation to cooperatives (Meynell 1984, 1990; Pollnac 1988b). Participation is a complex concept for which there is no universally applicable definition, partly because it can be interpreted as a graduated scale of involvement (Arnstein 1969). A crucial issue is whether participation is conceived as a means to an end, or an end in itself. 3.4.4.1  Participatory democracy  Although co-management can be a mechanism to manage consent and reduce conflict through participatory democracy (Pinkerton 1989), democratic co-management organizations can become the victims of power struggles and the general political environment (Jentoft 1989). The perspective on participation within an organization or country cannot be divorced from its politics. In the experience of non-industrialized countries, participation has generally been a project input of top-down development planning. It has been a short-term means to achieving project goals usually not formulated by the intended beneficiaries (Oakley and Marsden 1984). This token, manipulative or co-optational participation is incorporated for reasons of sharing project costs, increasing the efficiency of capital inputs and increasing the effectiveness of project outputs (Garcia-Zamor 1985). Since the failure of development projects is often attributed to sub-optimal participation, there has been a move in general development planning, parallel to the one in fishery management planning, to make planning more grassroots, bottom up or people-centred (Pollnac 1988b; BOBP 1990; Cernea 1991). In a country accustomed to development projects, but unused to fishery co-management, a pertinent conditioning factor may  47 be the extent to which the former was truly people-centred. 3.4.4.2  Empowerment  People-centred planning is an exercise in empowerment, and begs the related query about who are the ‘people’ involved. Here the co-management literature focuses on community development or participation in a ‘natural resource community’. This notion links the people who reside within some geographic area, determined to be relevant on various socio-spatial criteria, with their utilization of and cultural dependence on a particular renewable natural resource (Dyer and McGoodwin 1994). However this is only one interpretation of ‘community’, a term with numerous definitions. Social network research has shown that widely dispersed individuals can and do form communities because of their ties based on shared interests, values, norms or other commonalities (Wellman 1988). But if there are few ties between fisherfolk, if they are very individualistic, one cannot properly use the label ‘community’, a term typically invoking some sense of solidarity and mutual orientation (Scherer 1972). Since, in the literature, examples of the potential benefit of co-management to community development typically refer to a natural resource community (Berkes 1989; Pinkerton 1989; Dyer and McGoodwin 1994), broader application to other concepts of community within definition bounds is worth exploring. Social research also frequently elucidates a hierarchical power structure within communities and organizations, those who command the most social capital being the most powerfhl. In social network terms, the influence of powerful clients on the state, those who determine what is an ‘issue’ that gets dealt with, depends on the structure of the policy domain (Knoke and Laumann 1982). Since participation is about empowerment, particularly if it is an end in itself, there is the question of whether under co-management an existing elite within a community or organization will restructure the policy domain to become even more powerful or the mass of  48 people become newly empowered (Jentoft 1989). The latter is the intention of grassroots planning, but one that may be subverted by persons or organizations already in powerful positions who wish to defend their special status.  Delineating, through social research, the  networks connecting beneficiaries and the state may help to explain if and how social power could be redistributed and employed under a co-management scenario. 3.4.4.3  State and bureaucracy  However it should not be assumed that the state, specifically the government and bureaucracy, will want to relinquish and redistribute power for the common good. With the co management approach, adoption of participation as an end or means may require political change and the political will to institute bureaucratic reorientation (Montgomery 1986). Liberaldemocracy is not inherently antagonistic to participation, and may encourage it through interest groups where pluralist political beliefs are also prominent (Midgley 1986). What prevents politicians from adopting, and bureaucracy from reorienting towards, participative approaches may be more in the nature of particularistic barriers to change. Two major barriers, in terms of structurally altering networks of social capital, may be having to break both the patron-client relationships with fisherfolk beneficiaries, and the class association relationships with economic elites, from which power may be derived (Montgomery 1986; Pinkerton 1994). Socio-cultural and operational constraints to co-management may also exist for both the bureaucracy and the potential participants. These constraints may be most relevant to the decentralization  objective  of co-management  (Pinkerton  1989).  In  situations  where  subordination to authority, such as the state, is culturally embedded, it may be a pre-requisite to first raise the consciousness of the fisherfolk to a level where they have the desire and will to be involved in participatory processes (BOBP 1990). A fundamental requirement is that the fisherfolk voluntarily subordinate their private and short-term interests to the collective long-  49 term interests of a fishery organization (Jentoft 1989; Kuperan and Abdullah 1994). In individualistic societies such as described above for the Caribbean this may be difficult, but added to this are the typical operational difficulties associated with establishing effective organizations and mechanisms for interaction with the state. These include questions of technical and managerial skills, conflict resolution, communication and financial resources often related to the number and scale of organizations and the homogeneity of their membership (Jentoft 1989). This leads to the question of what, if any, role the state should play in institutionalizing participation. It has been argued that genuine participation can arise spontaneously only from the people themselves and that anything else is unsatisfactory and manipulative (Oakley and Marsden 1984; Midgley 1986). This view has been criticized as idealistically naive since the constraints described above are real and only through state assistance, sometimes termed bureaucratic populism, may participation be initiated successfully, eventually graduating to a more self-reliant arrangement (Montgomery 1986; Pomeroy 1994). In studying the role of fishermen’s organizations in fishery management around the world, Hannesson (1988) and Kurien (1988) generally conclude that, to be effective, these bodies need more recognition and support than they have been receiving. Whether such intervention is co-optation or manipulative is situation-specific, often determined by the degree to which the political and bureaucratic directorate are intent on neutralizing opposition and retaining or establishing control over the populace (Midgley 1986; Pinkerton 1994). The history of interaction between state and industry relevant to previous fisherfolk organizations can determine the response to state-aided initiation of co-management (Jentofi 1989). 3.4.4.4  Summary  In summary, the impetus to integrate social science into fishery planning and move towards co-management comes from the failure of the bioeconomic approach that ignores social factors  50  and the potential additional benefits of co-management. However problems exist both for the integration process and the institution of participatory planning. These problems have high political and social content and many are situation specific rather than generic. Thus the integration of social science may provide planners with the perspectives, concepts and methods necessary to collect and evaluate appropriate socioeconomic data useful for understanding the implications of social organization in the fishery. The case in the eastern Caribbean is that although there is uncertainty about the island states ever gaining a thorough knowledge of the pelagic resource, fishery development has been underway for some time. Since management planning is not yet practiced, but management and development planning are closely related, a review of the Caribbean development planning experience and environment may serve as a guide for what to expect in fisheries management planning. These factors are likely to be of relevance to co-management. The next section provides a general overview of development planning and administration in the eastern Caribbean before turning to the small body of literature on fishery planning. 3.4.5 Planning development in the Caribbean  3.4.5.1 Main features  Planning was introduced to the English-speaking eastern Caribbean in 1945 as a means of allocating Colonial Development and Welfare funds. From colonial to recent times, development planning has consisted mainly of thinly disguised listings of ongoing and desired high visibility, foreign-funded, capital projects in the public sector investment programme (Khan 1987). As a result of aggressive foreign borrowing, some countries have had to seek assistance from the International Monetary Fund (IMF), and planning is now done within the framework of liberal structural adjustment packages with less emphasis on capital projects. This situation has had several implications. Being small open economies, these states are  51 exceptionally dependent on the world economic environment, especially international finance and trade. It has been argued that true economic planning has never been done, nor is it possible, under conditions where there is no certainty of controlling the key variables to any appreciable degree (Barsotti 1992). With plans being primarily a collection of projects funded from different external sources, sectoral linkages have been poorly articulated and national sectoral input minimal (ISER 1982). Therefore, development planning, although supposed to achieve social ends through political and economic means, has been a mechanistic, technical process decoupled from the political process (Brown 1992). And although planning in the Caribbean is supposed to be indicative, neither plans, nor the political manifestos on which they are usually based, seem to be assimilated by either their authors or the electorate (Cox and Embree 1990). Human resources, in terms of adding the creativity of the citizens to the planning process, have been neglected (Khan 1987; Howard 1989). 3.4.5.2  Politics of planning  It is argued that the political education required for informed citizen participation has been lacking. “There is no serious attempt to bring the issues to the people, to give us a proper grasp of these issues and allow us to apply our wisdom to the process of problem-solving at the national level” (Hodge 1986:96). It is thought that political education should lay the basis for the greatest possible participation in planning to create a common understanding of what planning entails, and for policy to evolve from the people (Manley 1986). However, no political party in the Caribbean has shown enthusiasm for the political education essential for any major change in political consciousness since this would end the practice of personalized patronage politics (Wedderburn 1986; Khan 1987). The consequence of this political style is that small sections of the electorate seek direct access to politicians, bypassing the bureaucracy and supporting the tradition of political  52 patronage (Manley 1986). Large sections who are alienated from the political process express cynicism about the ability of governments to solve problems and work out their own strategies for survival (Hodge 1986). Plans are shelved if they constrain the freedom of action of the political directorate, and long term interests of a more general nature are subordinated to the short term interests of influential elite minorities adept at manipulating political structures and resources (Khan 1987). This means that long term plans and policy decisions are ignored and situation-specific decisions are taken for political expediency, an approach to planning described as reactionary (Cox and Embree 1990). Since the majority of people place little faith in democracy in the Caribbean, some writers see community councils and co-operatives to be essential if Caribbean states are to break out of non-democratic and non-participatory traditions (Thomas 1986). One purpose of such organizations would be to create, without political interference, greater commonality of various perceptions through information exchange and therefore greater capacity for common action in various situations (Manley 1986). The latter, which includes allowing more people to make inputs into planning, seems to share similarities with co-management. The above prescriptions, even though tendered from a socialist perspective, may perhaps still be viable in a true liberal-democracy. However, the states of the eastern Caribbean are said to be only nominally liberal-democratic. This ideology is allegedly threatened by the practice of governing which, to the present, is trapped within the colonial plantation value system of paternalism, authoritarianism and elitism (Stone 1985). These values, and the personalized political leadership style which causes the civil service to serve as a patronage vehicle rather than a change agent, places a limit on the degree of openness of government institutions (Khan 1987). Thus, popular participation is impeded by restricting the flow of and public access to information. Also, government agencies seem more concerned with control than client  53 satisfaction. When dealing with lower class clients who cannot fully articulate their needs, civil servants decide what is best for them using a top-down approach (Jones 1974). The tendency of civil servants to act on organizational premises and perceptions rather than those of the client is allegedly entrenched (Khan 1987). 3.4.5.3  Summary  To sum up, even today “one of the most significant characteristics of Commonwealth Caribbean systems of government and public administration is the non-participative nature of decision-making processes”(Jones and Mills 1976:333). These authors also say that when forced to facilitate participation these governments and bureaucracies engage in “symbolic institutionbuilding” (Jones and Mills 1976:341) so as to devise participative structures with no power. On Arnstein’s ladder (1969) participation in the eastern Caribbean is said to equate to tokenism, and it is therefore not surprising that citizen response to participative opportunities is often less than vigorous when the issues are not of direct interest (Khan 1987). However, at least one prominent regional public administration agency has recognized that the “broader dimension of public management is connected with empowerment of peoples through increased participation in the overall development process” and that there is a need for “bureaucratic reorientation” since the “participatory approach has acquired urgency as governments and the governed become more estranged” (CARICAD 1992:348). It is acknowledged that the popular image of the public service as self-serving, not nation-serving, noted by Khan (1987) needs to be improved. Yet formidable barriers to change may be set up by the political directorate rather than bureaucracy. As Hodge (1986:95) states somewhat cynically, political parties are usually intent on “keeping the people from making too many inroads on the resources which they control and intend to control forever.” This is not an encouraging environment for fishery planning based on co-management.  54 3.4.6 Progress in fishery planning  In the Lesser Antilles fishery planning has taken place within the environment described above from the inception of planning, but its part has been minor within the scheme of national development. Only since the mid-1980’s has fishery planning in the region been examined critically by national, regional and international agencies. By this time, academic researchers had already commented on deficiencies in the conceptualization and implementation of fisheries development plans (Kirton 1977; Berleant-Schiller 1981) but these comments either did not reach or receive the attention of those involved in planning. Gumy (1985), in reviewing the situation, observed that planning, although practiced widely, was not utilized in the fisheries sector as intensively as in the agricultural and industrial sectors. That is, fisheries development had either not been planned at all, or had not been properly planned. To make improvements, he and others (e.g., Cox and Embree 1990) have recommended interdisciplinarity, participation of the fishery sector, and greater harmonization between the actual decision-making power of state agencies and what is required according to the contents of plans (i.e., the reduction of political interference). However, all acknowledge that the technical, human and financial resources of state fishery agencies are inadequate in most countries. Under the auspices of the Food and Agriculture Organization (FAO) of the United Nations there have been follow-up activities. The first session of the Working Party on Fisheries Economics and Planning of the Western Central Atlantic Fishery Commission (WECAFC) held in 1989 again reviewed the situation and came up with general guidelines based primarily on conventional development economics with few social considerations (FAO 1991). Next, and more relevant, were the 1990 Workshops on the Economic and Social Aspects of Small-scale Fisheries in the Caribbean and on the Socio-economic Implications of Fisheries Management in  55  the WECAFC Region. The common conclusion from the workshops was that the typical bioeconomic approach could not deal with the “complexity of issues presently affecting the management of fishery resources [so] social and economic studies were considered as an important tool in the formulation and assessment of fishery management plans” (FAO 1991:3). The workshop also noted as a constraint the “lack of effective mechanisms enabling the active participation of fishermen in the formulation of management measures and often their reluctance to cooperate in the implementation and the accomplishment of management measures” (FAO 199 1:3). These statements were made at about the same time as an FAO Technical Cooperation Programme to devise fishery management options for the Lesser Antilles had concluded. Ironically, although the study noted the deficiencies in national and regional fisheries planning, and the need to get resource users involved in planning and management, the discussion of management options was almost exclusively biological (Mahon 1990a). However, in another regional meeting it was stated that “governments should establish mechanisms which will ensure regular and effective representation of all components of the fishery sector in the fishery decision-making process, for example Fishery Advisory Committees” (Chakalall 1992:13). The most recent follow-up is the CARICOM Fisheries Resource Assessment and Management Program (CFRAI\4P). This is a joint project between the Canadian International Development Agency (CIBA) and the Caribbean Community (CARICOM), a treaty-based political grouping of  English-speaking Caribbean territories. The 1991 Memorandum of  Agreement states that the goal of CFRAIVIP is “to promote the management of fisheries resources of CARICOM countries and to permit the exploitation of these on the basis of sustainable yield” (p.12). Therefore, included in CFRAMP is a sub-project on fisheries management planning.  56  The 1992 sub-project specification workshop (at which no fisherfolk were present) included discussion in favour of co-management and its associated features such as consideration of socioeconomic and political factors, fisher participation, more information exchange and evaluating fisher organizations’ potential role in management (Espeut 1992). But the bioeconomic fishery model appeared to dominate the technical sessions on management measures, with mention that ‘community involvement’ was useful in facilitating compliance (CFRA1VIP 1992). Community involvement was not a major and integral part of the discussions, and it does not appear that the issues involved in the bioeconomic versus co-management debate have been frilly articulated and evaluated. The fisheries management planning sub-project has not begun, and the latest strategic review of CFRAIVIP (1994) identifies the same issues as before. This is the point of departure for my research. 3.4.6.1  Summary  In summary, Caribbean fisheries planning has been subordinated to agriculture and located within a non-participatory development planning environment. However, within the last ten years this situation has changed to one in which fisherfoik participation is favourably discussed. Yet it is not clear which factors may determine whether the extent of this intended participation will be minimal as in the bioeconomic approach, or significant as in co-management. It is hoped that this study can make a practical contribution to the articulation and resolution of some of the issues in the case of the Barbados pelagic fishery and may also have broader application. A brief review of the analytical framework through which this may be accomplished based on the literature previously discussed is presented below in the final section. 3.4.7  Using this study as a planning resource The socioeconomic information gained from this study can be used by a fishery planner to  provide orientation for the interaction between the state and fishing industry. First, the  57 information could be used to evaluate the social organization of the fishery and select or modify the bioeconomic or co-management approach. Second, it could be used to plan the planning process  —  that is, the operational details of state/fishing industry interaction. Third, it could be  used in the execution of the planning process  that is, formulating, implementing, monitoring  and evaluating plan components. I illustrate some planning process options through diagrams showing the proportion of government versus fishing industry (a) participation and (b) decisionmaking in planning (Figure 3.4). The second chart illustrates that within co-management significantly different power arrangements are possible, and these need to be critically analyzed (Jentoft 1989; Pinkerton 1994). In the diagram, open access means that there is no management planning being undertaken by either party. This is the current state in the Caribbean to be addressed, so its continuation is not an option unless the situation proves to be unmanageable or management costs exceed benefits. Exclusive community management, in effect self-governance over a location or resource, implies that state intervention is and will be absent. Successful cases of communitybased management have been much documented, especially in reaction to the proposition by bioeconomists that some resources such as fish are by nature open access common property (McCay and Acheson 1987; Berkes 1989; Dyer and McGoodwin 1994), but examples from offshore fisheries are scarce (Jentoft 1989). Community management of migratory pelagics is not known to occur in the eastern Caribbean and so is unlikely to arise in the near future. Because of the nature of the fishery, a private property rather than communal solution to the commons dilemma has been proposed to be more suitable for Barbados (Berkes 1987).  58  Figure 3.4 Fishery p1aniing approaches  (a) Degree of state and fishing industry participation in the planning process max.  State participation  xmmwi propey  management  mm. Fishing industry participation  max.  (b) Degree of state and fishing industry decision-making in the planning process max.  State decision—malcing  common property mm. Fishing industry decision-making  max.  59  Co-optation means that the fishing industry takes no active part in decision-making, but is instead persuaded to implement state plans based on the belief that they are generally beneficial or inevitable (Kearney 1989). Two levels can be readily distinguished. However, in the most authoritarian or undemocratic situation the industry may not be consulted at all on the measures planned. In most countries where the bioeconomic approach is used the industry may be involved in the planning process to the point of being consulted for information or opinion, but no further (Jentoft 1989). I label these levels respectively co-optation (non-consultative) and consultation (consultative co-optation), and it is likely that consultation will be more feasible where there is some formal organization. Co-management was discussed previously and will not be reviewed again here. However, one of the anticipated benefits of co-management is reduced transaction costs because people buy into the plan (Pinkerton 1989; Berkes 1989). Yet, when successful, co-optation in either form assists in reducing the transaction costs of management even more than co-management by securing compliance without the price of fuller participation. But, as noted earlier, both levels of co-optation tend to fail for several reasons including lack of legitimacy, and hence compliance, among fisherfolk. Consider the first use of this study and recall previous discussion. If, for example, social strategies are found to be highly individualistic, the people used to authoritarianism, the potential to form effective fisherfolk organizations low, and the political and bureaucratic barriers formidable, then the bioeconomic assumptions and approach may be a feasible way forward, at least in the short term. In contrast, should the data indicate a high level of embeddedness with well developed social networks, effective fisherfolk organizations, and a desire for democratic participation by all, then conditions seem favourable to co-management. If the data present some  60 intermediate image then, depending on the degree of digression from the purest form of either approach, acceptance or modification of the closer one is an option. With the bioecononiic approach this may mean paying more attention to ‘imperfections’ due, for example, to kinship or unionization (Terkia et a!. 1988). With co-management, digression may mean careful nurturing and support of the elements that are favourable pre-conditions. The use of criteria such as discussed immediately above, and earlier in the literature review, facilitates an applicability analysis of alternative approaches. The second and third tasks of using the findings to plan the details of the planning process and execute it are closely related. Selecting the bioeconomic approach would suggest co optation or consultation, whereas co-management suggests participative planning. But in the latter, deciding which stakeholders participate, and how to share decision-making and responsibility between them, would still be at issue. If an intermediary situation exists, then the process chosen and the method of executing it would depend on the specifics of the social relations and organizations discovered. One method of assessing a suitable planning process and its execution is by examining which functions such as data collection and analysis, plan implementation, monitoring and evaluation, and enforcement are potentially sharable given the socioeconomic and political situation (Jentoft 1989; Pinkerton 1989; 1994). Comparing perceptions of uncertainty is relevant here since, as stated much earlier, shared perceptions form the basis of common action and problem-solving, while significant differences may form the basis for dialogue and information exchange, but also can be barriers to communication (Ruddle and Rondinelli 1983). Common interests, if not perceptions, could be expected to facilitate power and responsibility sharing in these functions. Perceptions of uncertainty also have practical implications for plan strategies, and examples  61 were given in the discussion of the use of information as a social power resource. However, the question of whether planners can actually influence perceptions of uncertainty among fisherfolk depends, among other things, on the networks that connect fisherfolk to different sources of information, and which of these sources they act upon. It may also depend on the quality and quantity of information available to planners. It is conceivable that, for example, fishers may have superior information usable by them as a power resource to gain leverage in orienting the planning process to their own needs (Ruddle 1994). Due to the paucity of data on the resources and fishing practices in the Caribbean, the functional expertise and perceptions of fisherfolk may be very valuable to planners (Berkes and Shaw 1986; Mahon 1 990b). Situations in which the state needs fishers’ experience and observations, and the fishers in exchange want more control over management, are not uncommon in co-management cases (Pinkerton 1989). Such a case allows relatively easy non-political entry into sharing power through data collection and analysis. Perceptions of uncertainty can also play an important part in determining the fundamental role and direction of fisheries management planning. On the subject of environmental management, Holling complains of the “fundamental myth” that “the central goal for design is to produce policies and developments which result in stable social, economic and environmental behaviour” (1978:2). Lamson (1984) observes that fisheries management planning has generally subscribed to this myth, and that it will be continually prone to failure unless reoriented for coping with variability and uncertainty rather than aimed at stability and prediction. She suggests that “a more appropriate set of goals would include: 1) learning how variability and fluctuation can be incorporated into resource policy design, and 2) seeking specific mechanisms that are capable of assisting individual fishermen and resource-dependent communities to survive unpredictable and sudden shocks” (1984:124). If one finds that uncertainty is perceived to be a pervasive and serious problem in the Barbados pelagic fishery, and accepts that the key method  62 for coping with uncertainty is flexible and adaptive fisheries management planning, then it may also be possible to choose whether bioeconomic or cooperative management planning is more appropriate based on socioeconomic information on the fishery. 34.7.1 Summary In summary, use of this study’s information involves evaluating: (1) whether the most feasible planning approach is bioeconomic, co-management or some intermediate; (2) the most feasible arrangements for participation and decision-making in the planning process; and (3) which, if any, planning functions are compatible with the anticipated level of state/industry interaction. These evaluations can be performed through the analysis of socioeconomic data on the social strategies that people in the industry use to cope with a pervasive feature such as uncertainty. These analyses involve comparing factors mentioned in the literature as positively or negatively affecting the above options with the actual situation based on interpreting the collected socioeconomic data. Before looking at the data collection methods used in my research, the situation in Barbados is briefly described.  63 4.  THE BARBADOS SITUATION  The purpose of this chapter is to provide the reader with an image of the study site through a brief historical and contemporary description. Many features of the fishing industry analysed in subsequent chapters are introduced here, but not elaborated upon. 4.1 COUNTRY PROFILE 4.1.1  Physical conditions  Barbados is the most eastern of the Caribbean islands, being entirely surrounded by the Atlantic Ocean and located at latitude 13° 10’ N and longitude 59° 35’ W. The mainly low relief and coraline island has a total land area of about 432 square kilometres encompassed by a coastline 95 kilometres long. The island shelf is small, only 320 square kilometres, and deep water is found close to shore. Since no marine boundaries have been negotiated, the potential extended marine jurisdiction has been estimated to cover around 48,800 square kilometres of ocean (Singh-Renton and Neilsen 1994). The surrounding oceanic surface waters are relatively low in nutrients, thermally stable and of low productivity. Surface currents off Barbados are complex but generally directed towards the northwest, sometimes bringing lenses of lower salinity water containing debris from the Amazon and Orinoco Rivers of South America (FAO 1993). 4.1.2  People and economy First settled by the British in 1625 and remaining under British rule throughout the colonial  period, Barbados became independent in  966. The Constitution of Barbados enshrines  parliamentary democracy based on the Westminster Model. In recent history the electoral contest has been between two major political parties with little difference in political ideology. Both describe themselves as social-democratic, but are typically described by others as essentially liberal-democratic (Stone 1985).  64  With an economy based for much of the past on sugar plantation capitalism and later agro commercial mercantilism, the growth of industrial capitalism has been slow. Post-independence economic development was based on Arthur Lewis’ Puerto Rico model of industrialization through attraction of substantial foreign capital investment (Howard 1989). The major focus is now on service industries such as tourism, data processing and offshore financial services. However, external and domestic factors caused negative growth in GDP during 1990, increasing deficits and rapidly depleting foreign reserves. This situation prompted the government in 1991 to enter into a stabilization programme with the International Monetary Fund (IlviF) aimed at reducing government intervention and strengthening the market economy, including free trade (Government of Barbados 1993). Table 4.1 shows the contribution to gross domestic product (GDP) at 1992 current prices in Barbados dollars for various industries. The same table shows the deployment of the labour force in 1992 when the official unemployment rate was reported at 23% of the labour force, mainly as a result of the stabilization programme, in an estimated national population of 262,600 persons. Overall female participation was 60.1% of the 1992 labour force. Table 4. iGross domestic product and employment in Barbados, 1992 Contribution to GDP at current prices (%) ($ Bds rnillions*)  Economic sector Trade and restaurants Government services Finance and business services Tourism Transport and communication  Manufacturing Agriculture and fishing Construction-related Otherservices Electricity, gas, water TOTAL *Exchange rate is 1 Barbados dollar Early elections were  498.9 480.4 453.1 317.3 248.6 203.2 163.8 118.4 107.7 100.6 2692.0 0.5 U.S. dollar  Employment of  labour force  (No.x 1000) (%) 14.2 14.7 18.6 17.8 21.7 20.9 16.8 4.1 4.3 9.6 10.0 11.8 9.2 4.4 4.2 7.6 10.4 10.0 6.1 6.2 6.0 4.4 7.7 7.4 18.8 4.0 18.1 3.7 1.8 1.7 100.0 100.0 96.2 Source: Barbados Statistical service  called in 1994, with the state of the economy and the relationship with  65  the I1VIF being central issues. The government changed hands, but one of the primary objectives typical of mainstream development economics, and stated in both the latest development plan of the former government and the political manifesto of the present government, is to restore the economy onto a path of “sustainable growth” (Government of Barbados 1993:42; Barbados Labour Party 1994:4). 4.2 THE FISifiNG INDUSTRY  Both administratively and statistically, fishing is included under the umbrella of agriculture. As shown in Table 4.1, and elaborated upon below, fishing is not a major contributor to the economy based on the official statistics. But, as FAO (1993) points out for the eastern Caribbean in general, the true value of the fishing industry is seldom accurately estimated due to deficiencies in available information. Some of the circumstances surrounding these uncertainties are introduced in this section. Table 4.2 chronicles some of the major events in the recent history of the fishing industry, not all of which will be discussed in the text. 4.2.1  Pelagic resources  Since the small island shelf cannot support a large demersal fishery, the multifleet, multispecies fishery for oceanic pelagics is predominant (Table 4.3). All of these species appear to be seasonal, some more than others, but in most cases the main season runs from November to July when over 90% of the annual catch is landed. Within the season there are usually peaks of abundance which shift from year to year and which are difficult to explain since so little is known about the fish, and all indices of abundance come from commercial catch and effort data (Mahon 1987, 1990b). Abundance indices based on such data can be exceptionally misleading even when the data are properly recorded using statistical design (Hilborn and Walters 1992), which is not the case in Barbados (McConney 1983; Willoughby et at. 1988).  66 Table 4.2 The recent history of the fishing industry in Barbados Year 1940 1942  .  1943  .  1944  .  .  • • 1945 1947  •  1949  •  •  • • • 1952  •  • • • 1955 1961  •  •  • 1962 1963 1964 1967  • • • •  • 1971 1974 1976 1980 1981 1982 1983 1985 1986 1991 1993 1994  • • • •  • •  • • •  • • •  Comment or event Fishing fleet consists of 371 sailboats and 165 rowboats employing 1200 fishers Report by H.H. Brown, the first comprehensive description of the industry, envisages a Fishery Department Fishing Boat Loan Scheme established and administered by the first Fisheries Advisory Committee Fishery Division established with grant funds under Colonial Development and Welfare provisions First Fishery Officer and one Clerk appointed A new Fisheries Advisory Committee chaired by the Director of Agriculture Fishing included in the first development plan Fishing Industry Act (FIA) passed which requires all boats to be inspected and registered with the Fisheries Division Colonial government begins to finance Fisheries Division after CD&W grant expires Tractor with winch installed at Tent Bay to replace manual labour for hauling up boats Fishery experimental boat “Investigator” launched Amendment to FIA requires fishers to be registered annually with Fisheries Division Gill net for flyingfish capture is introduced to fishers after 2 years of trials Boatyard set up at Fishery Experimental Station and “Calvert” design introduced Fisheries Division has a staff of 17 people Boat motorization started Hurricane Janet destroys many boats and ready supply of wood from felled trees facilitates boat rebuilding and motorization Fishing cooperatives introduced as savings societies Formation of the Barbados Marketing Corporation (BMC) The Barbados Fishing Vessels Co-operative Insurance Society formed but never active Workmen’s Compensation Act amended to include fishers as employees Most fishing savings societies converted into full cooperatives UNDP/FAO Fishery Project starts and introduces chilled fish to consumers through the BMC before its termination in 1972 Most fishing cooperatives have disbanded or become inactive Barbados Development Bank makes its first loan to the fisheries sector First small iceboat operated briefly First large iceboat is commercially operational Sand Pit fishers form informal group to counter relocation plans by the Barbados Port Authority First locally built fibreglass iceboat is commercially operational Some iceboat owners form a processing company, Barbados Fish Processing Ltd., to buy and market their catches Fish gluts caused by iceboats overwhelm the market Short-lived attempt to form a Fisherman’s Association The Barbados Fishing Cooperative Fishing Society Ltd. (BARFISHCOS), comprising mainly iceboat owners, is registered Barbados Union of Fishery Workers (BUFW) registered as a trade union A Fisheries Act, drafted to facilitate fisheries management planning, is passed Barbados United Fisherfolk Association (BUFFA) fonned  67 Table 4.3 Contribution of major pelagic species to estimated total landings  1986 1987 1988 1989 1990 1991 1992 1993 AVERAGE  All other species  Major pelagic species landed  Year Flyingfish  Dolphin  Kingfish  Tuna  Bilifish  (%)  (%)  (%)  (%)  (%)  62.9 36.4 65.3 54.4 58.4 52.7 44.0 69.7 55.5  21.6 55.3 22.1 29.6 30.8 34.5 43.7 18.0 32.0  3.3 2.4 3.6 2.9 1.9 2.2 1.5 1.9 2.5  1.7 1.2 2.6 3.2 2.4 2.6 3.3 3.5 2.6  3.3 1.7 3.7 3.2 2.8 2.1 2.5 2.1 2.7  (%) 7.2 3.0 2.7 6.7 3.7 5.9 4.8 4.9 4.9  Estimated total landings (metric tons) 4228 6696 9097 2754 2535 2074 3362 2852 4200  Source: Agricultural Planning Unit and Fisheries Division  Commercially, the most important of these is the small pelagic fourwing flyingfish (Hirundichthys affinis) which usually comprises about 55% of total annual landings. The first history of Barbados, written in 1657, marvels at their ‘flying’ (actually gliding) habit and mentions human consumption, suggesting that exploitation commenced shortly after settlement (Bair 1962). Flyingfish is of less, but growing, importance in the neighbouring islands and H. affinis appears to be the major species in all cases (Mahon et a!. 1986). The increasing regional capital investment in flyingfish capture, proceeding without knowledge of the potential yield, structure, or stability of flyingfish stocks, prompted the execution of a regional five year biological research project which started in 1987. The Eastern Caribbean Flyingfish Project had objectives of assessing the scope for developing the flyingfish fishery, investigating stock structure to determine the appropriate spatial boundaries for assessment and management, and developing feasible and appropriate management strategies (Oxenford ci’ a!. 1993). In addition to research dating back to the 1960’s, the project made flyingfish the most intensively studied of the pelagic species in the region. Many of the biological characteristics of H. affinis relevant to assessment and management have been described, but a  68 long list of required research still exists. Dolphin (Coryphaena hippurus, the fish, not the mammal) is the second most commercially important pelagic species in Barbados, usually comprising about 30% of the total annual landings. The biology of this large pelagic and its implications for the Barbados fishery have been studied, and hypotheses on stock structure and migration in the region have been put forward (Oxenford and Hunte 1984, 1985; Oxenford 1985). Both dolphin and flyingfish are schooling, short-lived (perhaps annual) species, and such stocks are notorious for providing little indication of overfishing before sudden and precipitous collapse (Hunte 1986). Besides sharks, which are usually not primary targets, the remaining pelagics of commercial importance are kingfish  —  chiefly wahoo (Acanthocybium solanderi), tunas  (Thunnus aibacares), bilifish  —  —  chiefly yellowfin  chiefly blue marlin (Makira nigricans) and Atlantic sailfish  (Istiophorus albicans), and swordfish (Xiphias gladius).  Hunte (1987) in reviewing the  available database on oceanic pelagics in the Lesser Antilles concludes that we are ignorant of the stock structures of most of the above and, except for the obvious overfishing of swordfish (Berkeley and Waugh 1986), the database is inadequate to conclude that these stocks are not being depleted. Little research has been done, or can comprehensively be done, on these fish in the eastern Caribbean since they are highly migratory and fall under the auspices of the International Commission for the Conservation of Atlantic Tunas (ICCAT). Through CFRAMP, fishery scientists and managers are gaining access to ICCAT data and attempting to encourage more research in this region where international fleets are known to fish, often illegally. These efforts have produced preliminary estimates of potential yields of large tunas and bilifish (Singh-Renton and Neilson 1994). This work is important to Barbados since it is thought that large pelagics, for which well-developed export markets already exist, hold the key to the future of fishery development in the region (FAO 1993). Although its 1990 landings  69 were only 0.3 9% of the total for the western central Atlantic (Stamatopoulos 1993), Barbados has already begun to invest in this direction. 4.2.2 Harvest sector  The Barbadian fishing fleet of over 500 boats is usually divided into four, not clearly definable, design categories. The smallest are open, oar or outboard powered dinghies (locally known as ‘moses’), less than 7 metres in length overall (LOA) used mainly for inshore capture of reef and slope demersals. These are not relevant to this study. The remaining categories (dayboats, iceboats and longliners) and their fishing methods are described below and in Figure 4.1. There is relatively little corporate or concentrated ownership, most owners being male individuals with a single boat (Table 4.4), and there are only about 3 female commercial fishers, one of whom is also the captain and owner. But it is commonly reported that only about 30% of the boats are owned by people who fish regularly (Berkes and Shaw 1986). While it is not clear how the composition of ownership has changed over time, it is generally felt that the proportion of fisher-owners has declined since motorization of the dayboat fleet in the 1950’s due to the increasing capital requirements of initial investment and operation (Table 4.5). Certainly fewer fishers than non-fishers own iceboats or longliners. Dayboats are typically decked wooden launches 7 to 11 metres LOA, propelled by inboard diesel engines, which carry two fishers and land their catch daily since they carry no ice for the preservation of the catch at sea. On the same trip a dayboat may attempt to catch flyingflsh by gillnet with screeler (a tethered fish aggregating device) and handline, and any or all of the large pelagics by trolling or drifting (lurk) handlines. They normally range up to 20 miles from shore and there were about 370 dayboats in 1993. A typical dayboat lands 6.5 metric tons of flyingfish and 3.5 metric tons of large pelagics per year (Hunte and Oxenford 1986).  70 Figure 4.1 Fishing boats and methods a) Dayboat using screeler to drift giinet for flyingfish while also handlining (lurklining) for large pelagics  b) Iceboat trolling for large pelagics  c) Longliner setting for tuna and bilifish  Iceboats can be very similar to dayboats except for size (10 to 15 metres LOA and 3 fishers). They carry ice to sea in insulated fish holds, facilitating trips of several days duration, and usually harvest the same species as dayboats using the same gear, but sometimes more of it. The important difference is that they range further in search of fish, often up to 200 miles or more. In 1993 there were about 76 iceboats. A typical iceboat lands 17 metric tons of flyingfish and 9.2 metric tons of large pelagics per year (Hunte and Oxenford 1986).  71  Table 4.4 Boat ownership by gender, concentration and type Iceboats and longliners  Dayboats  Ownership criterion  (%)  (%) A. Gender male female B. Concentration ownslboat owns 2 boats owns 3 boats owns 4 boats C. Type sole owner partnership  (N340)  (N82) 91.5 8.5  92.4 7.6 (N82)  (N=340) 91.8 7.9 0.0 0.3  85.4 11.0 2.4 1.2 (N=87)  (N=370) 90.5 9.5  87.4 12.6  Source: Fisheries Division Nov. 1993 boat registry database  The 1993 longliner fleet consisted of about 11 boats varying in length from 12 to 30 metres LOA. They stay at sea for 2 to 3 weeks and range sometimes more than 400 miles. With 4 or 5 fishers on board they use the gear and methods of Florida-style vessels. Their target species are tuna and swordfish for export, with bycatches of shark and billfish sold locally. Due to the greater size and fishing method variations among longliners it is difficult to assign a typical year’s catch, but over 50 metric tons of total pelagics is not unreasonable (Bellairs Research Institute 1994). In the early 1980’s considerable conflict between the dayboat and iceboat fleets was caused by frequent localized fish gluts. These gluts, characterized by record low prices for both fleets and dumping at sea of excess iceboat fish, resulted from the rapid state-financed expansion of the iceboat fleet in conjunction with an inadequate fish marketing and distribution system. It was said then that iceboats had the potential to drive both themselves and dayboats out of business unless fleet growth slowed or the postharvest situation improved (Lavine 1984). Due to input  72 and infrastructure constraints on iceboat operations and fleet expansion, this competition and conflict with dayboats is not now as apparent although it still exists. Since most of the longline landings do not compete on the local market they seem not to have been drawn into the fray. However, no statistics show the contributions of the three boat types to total landings or earnings, and the direction of future fleet development is not clear since recent tough economic conditions in the country have prompted a renewed interest in smaller, more versatile, boats.  The numbers in Table 4.5 should only be taken as a very rough guide to vessel economics. No financial data from complete, detailed actual expense and revenue records are available for any of the boat types since few owners keep or divulge such records. From the 1940’s to the present, reports that have claimed to show the profitability of “typical” boats have varied so much in their conclusions as to whether boats are making profits or losses, and the amounts involved, that the true situation is still a matter of speculation (McConney 1987). The table depicts only one of several scenarios, and it is relative rather than absolute values that deserve most attention. Added to the uncertainty of the financial data are the several different ways that accountants, economists and others have of calculating profitability. For example, while the increase in overall enterprise profitability and fuel efficiency with scale is generally agreed upon, profit per ton of fish is lower for an iceboat than dayboat, while investment per fishing job created is much higher (Berkes and Shaw 1986). Whether increasing scale is good or bad depends on one’s perspective. The situation with personal cash income to owners and fishers is similarly subject to interpretation.  73  Table 4.5 Estimated fishing income and expenses Item  Dayboat  a  Iceboat a  ($)  Longliner  b  ($)  ($)  375,000 45.000 200,000 Capital cost Annual enterprise profitability 34.540 90,640 1,003,226 Total revenue 23.702 948,704 65,726 Total costs ° 24,914 54,522 10,838 Net revenue Personal annual net cash income 24,522 7,076 11,640 Owner/captain d 10,638 10,000 6,764 Crewman 1 10,000 10,638 Crewman 2 10,000 Crewman 3 Derived from: (a) Hunte and Oxenford (1986); (b) Bellairs Research Institute (1994) (c) includes loan, insurance, maintenance, operational and depreciation expenses (d) net of loan, insurance and maintenance expenses —  —  —  If the dayboat owner is the captain, he or she gets 75% (i.e., fisher plus boat share) of the gross revenue minus operational expenses, and the single crew gets 25% (Hunte and Oxenford 1986). They report an iceboat owner/captain getting 70%, with 15% to each of the two crew, and say that these share systems are fairly typical. For both boat types, when you subtract the owner’s cash expenses (i.e., excluding depreciation) there is little income difference between an owner and fisher in the short-term. In the long-term for an owner/captain, and in both the short and long-term for non-fishing owners, income tends to erode to a level below that of fishers. As discussed in Chapter Seven, fishers seem not to realize that an owner’s returns to equity can be negative (Burtonboy and Horemans 1988), and the only financial incentive for owning a fishing boat appears to be short-term cash flow when fishing is good. A typical picture for longliners is even more difficult to present due to the vessel variability mentioned above, and the personal incomes represent only base estimates. Several different share and other payment systems are in use as owners experiment with incentives to retain skilled fishers and increase their productivity.  74 Due to the increasing capitalization of the fleet and the competition for fish at increasing fishing ranges, it has been reported that Barbados is experiencing a tragedy of the commons situation (Berkes 1987). It is likely, however, that deficiencies in the postharvest sector have been important in inhibiting the hunt for that last fish. 4.2.3  Postharvest and supporting activities Pelagic catches are landed mainly at the sites shown in Figure 4.2 and sold there by one of  the fishers, the boat owner or an agent. Ex-vessel sales approximate to brief informal auctions depending on the species of fish, the amount offered for sale and the relationship between sellers and buyers. The buyers are usually vendors (hawkers) in the case of dayboats and some iceboats, or small processing companies in the case of many iceboats and longliners, particularly when landings are large. At present there are 5 fish processing plants, varying greatly in scale, but each easily identifiable with a particular individual rather than being anonymously corporate. There may be about 200 regular fish vendors, but the actual number operating varies with the volume of landings since the occupation is opportunistic and many more people sell fish when it is plentiful. During the season some vendors buy for processors in addition to their own trade, and in the offseason buy imported or local stockpiled frozen fish from them for resale to consumers. Overall, most of the vendors are women, but this varies with location. Most vendors and processors handle all species of pelagic fish except for swordfish which is the preserve of processors. Fish is sold directly to consumers at the landing site by vendors, and to supermarkets, restaurants and hotels by both vendors and processors. At all sites some fish is sold by the fisher or owner to buyers other than vendors or processors, but this varies by location and volume of landings. Figure 4.3 shows some typical distribution channels.  75 Figure 4.2 Fish landing sites  LANDING SITES • Primary  • Secondary Tertiar  Moon Fcfl  !  Si) I’Ien’s htsfnxvn  ATLANTIC OCEAN  ‘s Bay Tent Bay  °“nes Bay  Conset Ba  Bay Pit  Approx. 8 km  13°OO’ N 59°40’ W  A number of other occupations, not directly part of this study also play supporting roles. For example, many vendors make use of ‘boners’ and ‘scalers’ to assist them in processing whole flyingfish into the boneless butterfly fillets increasingly demanded by consumers for convenience. The relationship to vendors varies from being virtual employees, working exclusively with one person and sometimes assisting in sales, to being for general hire by anyone  76 including consumers. At a time of high unemployment, this form of self-employment with its minimal capital outlay has attracted both young men and women, thereby increasing specialization and the gender neutral division of labour in the industry. These jobs have the drawback of being lucrative only when landings of flyingfish are large. Similar observations apply to ‘dolphin skinners’, another specialty task, but these are almost exclusively male and concentrated at Oistins where dolphin landings are greatest. Figure 4.3 Simplified fish marketing and distribution channels  Boat agents or fish sellers (not vendors) are typically older men operating at Bridgetown. On behalf of dayboat owners, once the fish is landed they sell to vendors or the general public at the best price in order to receive a commission. Many pay the fishers their share and hold the profit for the owner to collect later. Some agents are also vendors. Another category of fish  77 seller is the man who trucks fish from smaller landing sites to Bridgetown, sometimes selling fish along the way. This added expense to boat owners is necessary since market demand is insufficient at these sites. Some fish transporters are essentially ‘speculators’, buying fish cheap anywhere and selling where prices are higher. Boatbuilding is another major supporting activity not included in this study. The wooden boats in all categories are locally constructed from imported materials mainly by a handful of individual artisanal shipwrights using (without blueprints) a traditional design that has changed little over the past 40 years. Fibreglass boats are fewer and are imported or built locally by two commercial firms. Very recently a new boatbuilding division of labour, the fibreglassing of wooden boats, has arisen. Finally, there is a general category of person, usually male, who performs almost any task, but typically assists in unloading and washing down the larger boats, runs various errands for payment in kind or cash, and sometimes becomes a crewman if the opportunity arises. These people are found mainly at Bridgetown and Oistins. Most of the fish landed are consumed locally by Barbadians and visitors. Besides the demand from tourism for types of seafood that the local fleet does not supply, Barbados also imports quantities of fish mostly in the offseason to top up supplies for general consumption. The expected inverse relationship between local production and imports is obscured by a number of factors including doubt about the statistics on the amount of fish landed. The most commonly reported total annual landings for the past 30 years have been between 3,000 to 5,000 metric tons with lows and highs to nearly 2,000 and 10,000 metric tons respectively, depending on the source consulted. For the same period, annual fish consumption has been estimated at 25-30 kilogrammes per person or 20 % of animal protein supply (Laureti 1992). Local fish is seldom processed beyond being frozen and filleted or steaked, and must compete with both imported  78 value-added convenience seafood and lower-priced basic items. Bearing in mind the warning about official estimations of GDP given earlier, the low contribution of fishing in Barbados is apparent from Table 4.6 which sets out landings, trade and economic statistics. Table 4.6 Estimated total fish landings, contribution to GDP and trade Year  1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993  Estimated total fish landings a (let C tons) 3735 (or 3672) 3411(or3963) 3480(or 3435) 6522(or 5904) 5787 3915 4227(or2956or9200) 3702(or9800or6696) 9097(or5300or7800) 2547(orS000or2O5l) 2967(or2535) 2698 3283(or3341) 2852  Fish imports Contribution to GDP at live weight equivalent current prices b 0.7 0.6 0.6 1.0 0.9 0.9 0.9 1.0 0.8 0.6 0.9 0.6 0.8 0.8  3044 4943 4164 3010 3527 3993 3744 3765 3461 4646 4747 na na na  Fish exports live weight equivalent (mçtric 40 36 14 35 26 22 31 28 182 69 61 na na na  a = FAO FISHSTAT (or Economic Report or Agricultural Planning Unit) b = Barbados Statistical Service c = Laureti (1992)  4.2.4 Administration and organizations  As stated earlier, the pelagic fishery is not managed in the sense of being regulated for conservation or utilization. It has, however, been developed largely through state intervention in the form of physical infrastructure at landing sites, and provision of credit for boat construction. Fishing as a subsector of agriculture has been mentioned in all development plans since the first in 1945. It had received attention before that through the appointment of a Fishery Advisory Committee (FAC) and Fishery Officer (see Table 4.2). However, in comparing the contents of development plans to what actually took place, one can conclude that fishery development has essentially been ad hoc (McConney 1987).  79 Except for brief periods, responsibility for the fishing industry, and most of the government agencies directly concerned with it, has been with the colonial or government body responsible for agriculture. The more indirect services of credit and insurance have been provided, postindependence, by statutory corporations under the finance ministry. The only agency whose jurisdiction is exclusively the fishing industry is the Fisheries Division. The Fisheries Division, which originated with one Fishery Officer in 1944, now employs about 25 persons (including one Fisheries Biologist) most of whom are engaged in providing services mainly to the harvest sector. Some of these services such as free engine maintenance, a variety of subsidies and boat haul-out by tractor are diminishing through attrition or government’s recent structural reform policy. Other functions such as boat registration, boat inspection for safety, training and research of all sorts, statistical record-keeping and fisheries management planning are being implemented within severe operational constraints including a small operating budget. These functions are linked in part to a new Fisheries Act (1993) which gives the Fisheries Division responsibility for fishery management and planning. The latter includes formal consultative mechanisms through which persons in the fishing industry can advise the minister on a wide variety of fishery-related topics. The operational constraints are argued to reflect the low status of the fishing industry and Division among government’s priorities, except for capital works (Research and Productivity Council 1980; FAO/IC 1982). The Fisheries Division’s status in the parent ministry, especially in relation to planning, is discussed in the penultimate chapter. The Markets Division is responsible for operating the state-owned premises where agricultural produce is sold to the public. These include the three fish markets at Speightstown, Oistins and Bridgetown. The latter, costing twenty million dollars in 1989, consists of fishing harbour, boat repair yard, ice production, chill and cold storage, blast freezing, processing hail  80 and retail areas. It is the major fish landing site, serving up to 150 boats. Oistins, costing five million dollars in 1983 and serving about 80 boats, is the second largest site consisting ofajetty and scaled-down versions of most of Bridgetown’ s facilities, except blast freezing. The Markets Division, unlike the Fisheries Division, has no development mandate and is more concerned with post-harvest services. It is also responsible for recording daily fish landings and prices at the fish markets. The Agricultural Planning Unit is also concerned with fisheries statistics, but in collating not collecting them, except for the occasional survey. Its relevance to this study is its primary role in writing the fisheries section of the agricultural sector development plan which subsequently becomes incorporated into the national development plan. Since fishery planning is located entirely within the agriculture ministry, the national planning process is pertinent only in terms of setting guidelines for the sectoral exercise and editing the output. Such guidelines set up a consultation system to facilitate input, by the private sector and ‘technical’ departments such as the Fisheries Division, into the latest development plan. Other former arms of the agriculture ministry were important in shaping the fishing industry in its early stages. The Cooperatives Division, which introduced co-ops to the fishing industry in the early 1960’s, some would argue was also partly responsible for their sharp decline within the decade (Wedderburn 1981). This negative experience with producer co-ops in general, and fishing co-ops in particular, apparently shapes fishers’ current attitudes towards grassroots collective action despite the considerable success of credit unions. Pre-independence political ideology played a critical role in the promotion of cooperatives, and the intention of promoting fishing co-ops still appears in every plan. This is despite lack of action, and in the face of evidence that this may be a lost cause (Burtonboy 1988). Collective action is considered in detail later.  81 The Barbados Marketing Corporation (BMC), a statutory body formed in the early 1960’s, was also part of the political philosophy of the day. It offered development support by working with fishing co-ops and offering a guaranteed market and prices for fish. Subsequently, for political, economic and technical reasons, the BMC’s fisheries operations were terminated, but not before it had shaped expectations of how the government should intervene to make life easier for the harvest sector. Some of these expectations, which remain alive today in an inhospitable political and economic environment, are discussed later. Credit for fishing fleet modernization and expansion originally came from the colonial Fisheries Division whose revolving loan fund ground to a halt decades later due to low repayment. Now credit is provided mainly by the Barbados Development Bank (BDB). Apparently without the benefit of detailed technical or socioeconomic evaluation, the Bank was instrumental in facilitating the iceboat fleet expansion that changed the face of the industry in the 1980’s (McConney 1987). Under pressure from international donors to reduce its nonperforming loan portfolio, the BDB is in the process of restructuring, and its future relationship with the fishing industry is uncertain (W. Lavine pers. comm.). The BDB is relevant to this study due to its lack of formal connection to the other agencies, and the fact that its credit policy is a major factor in shaping investment in the harvest sector. This is of concern in fisheries management planning. The Insurance Corporation of Barbados (ICB) is the main insurer for the fishing fleet, especially boats financed by the BDB. It is relevant only in that objection to its allegedly high premiums may be one common basis for collective action. Regarding collective action, there are currently no fully functional organizations in the fishing industry. As shown in Table 4.2, after the failure of the early cooperatives there were short-lived attempts at collective action. These included a purchasing and processing company operated by iceboat owners, a fisherman’s association, a cooperative of mainly iceboat owners, a  82 general fishery worker trade union, and an embryonic association of mainly boat owners and fishers. All of these are reviewed later, but especially the latter which was formed as a response to uncertainty and provided an opportunity for participant observation. This and the other research methods used are discussed in the next chapter.  83  5.  RESEARCH DESIGN AND METHODS  In this chapter, I begin with the research design, describing why various methods were chosen and how they were combined to answer the research questions. The methods used to collect and analyse the data are then each explained in detail. This detail is useffil from an applied perspective since research on the Barbados fishing industry by state agencies and the local university is becoming more common, but information on methods is seldom exchanged to facilitate learning from experiences. A brief discussion of the methods used concludes the chapter. 5.1 RESEARCH DESIGN 5.1.1  General considerations Research, from formulation to conclusion, implicitly involves making choices about  alternative designs at each stage. The benefit from making design choices explicit is that you can better determine the appropriateness of design at each stage and the applicability of the final results. General considerations like the nature of the study, the population of interest, and methods are the pre-sampling choices of research design (Henry 1990). To reiterate, the choice of research problem was based on the desire to apply interdisciplinary social science to a real situation. As a consequence of the lack of information on the specific circumstances in Barbados, the study is largely exploratory and data-driven. But it is also analytical in terms of investigating relationships, and prescriptive in terms of recommending alternative courses of action based on the findings. It would be ideal for the purpose of theory-building to generalize the research results at least to the entire small-scale pelagic fishery in Barbados and preferably to the wider Caribbean or elsewhere. But it was known from the outset that the statistical requirements for external  84  validity could not be fulfilled. Yet it was intended that the methods chosen should provide results that were indicative, if not representative, of features within the pelagic fishery, Therefore some thought went into selecting combinations of methods that, if used together, would justify the claim that results were more than particularistic. Considerations such as limited time and financial resources, the physical and socioeconomic aspects of the location, the lack of earlier research on which to build, and the anticipated application of the findings also shaped the choices of research methods. In all cases complementary methods were brought to bear on each question as shown below. 5.1.2 Perceptions of uncertainty  The first research question asks: “What uncertainties are perceived to characterize the Barbados pelagic fishery?” It is not the intent to investigate the causes of perceptions, but only to describe the perceptions themselves, and this was most feasible through self-report by means of structured, semi-structured and unstructured interviews. Since, similarities and differences in perceptions have consequences for common action and dialogue, respondents both in the fishing industry and in government were interviewed using some common questions. The uncertainties investigated were those most frequently reported in the general literature, but especially those which the documentation available on Barbados also addressed. This was done in order to compare perceptions and measurement of uncertain phenomena, and to identify areas of possible collaboration or conflict between fishery officials and industry participants. The uncertainties associated with some of the government records, and the extent of available planning information, were also identified through document study and interview. Also, by listening to, and participating in, conversations and activities at fish landing sites it was possible to get indications about what aspects of their work experience people in the fishing industry felt were most uncertain, and what could be done about them.  85 5.1.3 Social strategies for coping  The second research question asks: “Why are particular social strategies used to cope with these uncertainties?” As noted above, some of this information came from participant observation at fish landing sites. People said whether they dealt with things by themselves, through relations with others or if they thought formal organization would be the solution. It was possible to directly observe interactions that could be interpreted as examples of network relations, particularly with assistance from key informants. The opportunity also arose to apply participant observation to the initial stages of the formation of a fisherfolk organization in response to a perceived uncertainty. This was almost akin to ethnographic research. The detailed investigation of social strategies for coping was carried out later in the research. Notes from the participant observation and unstructured interviews at fish landing sites, and the analysis of the data from the preceding fishing industry questionnaire survey, allowed design to be both theory- and data-driven, that is, partially based on prior results. Thus the social network research related to coping with the uncertainties perceived by fisherfolk to be most important, and the research on formal organizations by semi-structured interviews related to those organizations most discussed, and the persons most involved. The state intervention aspect of fisherfolk organization was also studied by examination of official documents and files in order to fill in the historical gaps for periods outside the accurate recall of respondents, and for which no other records exist. This overlapped with consideration of planning. 5.1.4 Implications for planning  The third research question asks: “What potential implications do these perceived uncertainties and coping strategies have for fishery planning?” Questions in structured and semi structured interviews soliciting opinions on various aspects of planning were posed to both  86  fishing industry and government respondents, but mainly the latter since they are the ones expected to apply the research recommendations. Since there was little evidence of state/industry planning interaction available for direct observation, self-report was the major tool. Comparison of the opinions of officials, industry participants and the policy statements in documents highlight areas of potential conflict and cooperation within the bureaucracy, and between the bureaucracy and industry However, much was also made of documentation describing both the historical and contemporary planning and political environment. The timing of national elections during the research period facilitated the latter investigation. The legislative basis and administrative preparedness for fisheries management planning was also included in the document study. 5.2 RESEARCH METHODS 5.2.1 Survey of fishing industry  This survey, code-named QFISH, was primarily exploratory, the main purpose being to provide data useful for compiling a general picture of the pelagic fishing industry in terms of perceptions of uncertainty, social coping strategies and opinions on planning. The data was partially analysed in the field to inform the design of the subsequent research methods. The requirement was to rapidly survey a large number of people in the fishing industry on several of the areas described in the literature review. Although the cost per respondent is usually lower for phone or mail surveys, these options are not feasible in Barbados, especially for the fishing industry, since adequate sampling frames do not exist. These methods would also have been inappropriate since the benefits of personal interview such as higher response rate, greater spontaneity and flexibility, and the ability to observe non-verbal behaviour outweighed the disadvantages of the method (Bailey 1982). Guidelines for social survey design, administration and data processing (e.g., Converse and  87 Presser 1986; Rea and Parker 1992) were consulted. Also reviewed were special considerations for developing countries (Bulmer and Warwick 1983) and the Caribbean (Barrow 1983). 5.2.1.1  Respondents  In QFISH the study population comprised the principal players in Barbados’ and most small-scale fishing industries: fishers, boat owners, fish vendors and fish processors. However, it was not possible to rely on random techniques to select interview respondents. The practical problems of random sampling in small-scale fisheries surveys are well documented and have led to the use of non-probability samples in Barbados and elsewhere in the Caribbean (Poggie 1979; McConney 1987). The fundamental problem is that adequate sampling frames are not available. However, even if they were available, fishers and fish vendors (to lesser extent) are difficult to physically contact on the basis of publicly available records. Many will not have telephones or current addresses listed, they are usually known by nicknames rather than the names on official records, and they work long and irregular hours often at several sites. Therefore to identify and locate randomly selected respondents is prohibitively time-consuming, if not impossible. Although inadequate for random sampling, registration and licensing lists of boat owners and fish vendors are available from the Ministry of Agriculture, Food and Fisheries (MAFF). There are no lists available for fishers, but the few processing plants are in the telephone directory. Using these lists, minimum populations of the respondent categories were estimated in order to draw quota samples reasonably representative of their numerical proportions in the industry. Fish processors were the exception since, with only 5 operating, a census was feasible. The means of calculating quotas are described below. Errors and omissions in the official records are highlighted since these are some of the uncertainties which I do not describe again in detail, but are the realities fishery planners now face. Since persons who sell fish to the public are required by law to be licensed, the best  88  available list of fish vendors is the Register of Fish Seller’s Licences maintained by the Markets Division. However, some sellers within, and most sellers outside, the environs of the public markets may not be licensed, and the number of unlicensed vendors varies during the season. Many people, including specially recruited relatives and friends of vendors, sell fish opportunistically during periods of heavy landings and when alternative opportunities for employment are few. Rough estimates of the numbers of licensed and unlicensed regular fish vendors were obtained from observation and inquiry and, it is reasonable to conservatively estimate the minimum population of fish vendors at 200 people. For the harvest sector, a computerized database of owners and registered fishing boats, expected to have been relatively accurate for the 1992-93 fishing season, was obtained from the Fisheries Division. The numbers of commercial fishing boats and their owners in mid-1993 were as in Table 5.1. However, one must take into account possible errors of both over and under estimation. Overestimation occurs since not all boats listed may be actively fishing due to repairs, dereliction or wreckage. Underestimation, probably the greater error (E. Holder pers. comm.), occurs due to an unknown number of boats that are unregistered, but believed to still be fishing, and boats recently registered which are not yet on the database. Table 5.1 Harvest sector populations Boats Owners Boat type 370 340 dayboat 76 74 iceboat 11 longliner 8 457 422 TOTAL Source: Fisheries Division records  Fishers per boat 2 3 4  Total fishers 740 228 44 1012  Since the database does not contain information on the occupations of boat owners, it is not possible to determine how many are also fishers, vendors or processors (i.e., occupy dual respondent categories). Yet it is believed in the fishing industry and the Fisheries Division that  89 the majority of boat owners (70-80%) do not regularly fish on their boats, and the frequency of ownership combined with other fishery occupations is low. Assuming that the register underestimated the active fleet at the time of the survey by about 20%, then it is reasonable to estimate the population of non-fishing boat owners at 422. Boat owners who fish are included among the fishers for the purpose of sampling. The Fisheries Division estimated that 1,875 fishers were active in 1993 in all of the fisheries, but since there is no sampling frame for pelagic fishers, an estimate of their minimum number was obtained by multiplying the numbers of dayboats, iceboats and longliners by the official records of the typical complement of fishers onboard. The calculations are shown in Table 5.1. The grand total represents the estimated minimum number of fishers that would be on all the boats fishing at the same time. Errors are carried over from the estimation of the number of boats in the fleet, and are introduced by the actual crew complement being different from the official records in some cases, and the fact that since all boats do not fish simultaneously, fishers frequently shift between boats. However, it is not unreasonable to expect that the number of pelagic fishers will be close to 1,012. It is also logical to assume that the number of captains equals the number of boats, but this neglects the practice of captaincy rotating fairly freely, especially in the dayboat fleet, between the regular captain and the ‘second man’ when the former is unavailable. As a consequence, no distinction was made between captain and crew in assigning quotas of fishers. Based on available time and financial resources, the researcher being the sole interviewer, and the exploratory nature of the survey, it was expected that about 200 interviews comprising samples of fishers, boat owners and vendors and a census of the processors would be feasible and adequate. The populations of respondents used for this study as estimated above, and the interview quotas assigned proportionately to respondent category are in Table 5.2  90 Table 5.2 Quota allocations and actual interviews executed Category vendor owner fisher SUB-TOTAL processor TOTAL  Population os) 200 422 1012 1634 5 1639  Population 12 26 62 census  Respondent 23 51 121 195 5 200  Respondents intrv (nos.) 32 40 126 198 5 203  Given the very crude nature of estimating populations, the quotas were intended to be guides only, to be modified in the field according to the numbers of persons in the various categories actually observed. Quotas could most easily be filled by approaching potential respondents at fish landing sites where they were working or relaxing. 5.2.1.2  Sampling sites The Fisheries Division recognizes about 30 fish landing sites. Pelagic fish are landed at  approximately 15 of these during the fishing season. For logistical (time and travel) reasons sampling was done only at the 9 most active sites. Based on the number of boats listing them as their normal operating bases, their recorded fish landings for the 1992/93 fishing season, the numbers of fish vendors and fishers estimated to operate at them, and observations on preparedness for the current fishing season, each of the sites was subjectively weighted. The weights were Bridgetown (xl 5), Oistins (x8), Sixmens (x2), and Speightstown, Pile Bay, Reads Bay, Skeete’s Bay, Conset Bay and Half Moon Fort each (xl).This exercise was done to allow quota sampling to proceed on the basis of landing sites being visited on randomly selected days, with the probability of site selection weighted by the anticipated relative importance of the site. The purpose was to reduce potential bias from selecting visits purely on the basis of convenience. Except for Bridgetown and Oistins which are active all day, the sample sites were usually visited in the afternoon between 14:00 and 18:00 when fish landing and marketing is at its daily peak. To further reduce bias, transects were walked in random directions across the site,  91 and anyone not obviously engaged in an activity which could hinder the interview was approached, the study explained and consent to interview sought. The questionnaire was read aloud to the respondent and filled in by the researcher. This was the most feasible arrangement which also avoided complications due to the possibility of respondent illiteracy. 5.2.1.3  Instrument  My requirement was for an interview schedule that could be administered in about 20 minutes given the number that had to be completed by the researcher and the need to minimize the imposition on respondents. The questionnaire needed to cover the following areas: • Respondent attributes such as work category, gender, age, education; • Perceptions of uncertainties in the environment, resource, operations, economics; • Social relations involved in coping strategies such as information exchange, marketing arrangements and credit provision; and • Opinions on fishery planning such as what objectives are important, who should work towards them and the nature of government/industry interaction. The QFISH questionnaire in Appendix A is a result of modifications made following pretests. The pretests were mainly to assess the interpretation of the concepts, clarity of responses, task difficulty and level of interest among respondent with various attributes. The need for a second pretest was based on the results of the first which showed ftirther refinement was necessary. To start the study, and test the phrasing of questions in the everyday language of fisherfolk, I first had informal interviews with people at fish landing sites who were not aware of the impending survey. The instrument was then tested on 3 people in the fishing industry (fisher captain, fisher crewman, female fish vendor) conveniently selected from among persons who appeared to be typical of their occupations at the main fish landing site. They answered the  92 questions on the interview schedule under typical field conditions, and then were asked to give their views on the interview experience. I noted separately any difficulty in responding to questions and found that respondents did not seek clarification or explanation, but were providing minimal and irrelevant answers, where they were confused. In another Caribbean survey Barrow (1983) warns about over-estimating respondent comprehension. Revisions were made and the second draft schedule was tested at fish landing sites on 20 respondents purposely selected to include a range of persons in each respondent category. The main selection criteria were various lengths of experience in the industry, and variations in anticipated ability to comprehend the questions (based on previous casual conversations). Again, at the end of the interview, additional questions were asked in order to obtain feedback on the experience, and separate observations were made by the researcher. These results were more positive. I found that, for example, the background variables were not problematic except that accurate and reliable responses to queries on absolute income were not forthcoming, so a question soliciting relative income was substituted. This pretest also showed that the notion of things in the industry being uncertain (unknown or unpredictable) was a familiar one. This was important in order to validate one of the main concepts. But people could not easily scale their perceptions of uncertainty. With few exceptions they thought that things either were, or were not, usually predictable. Hence most uncertainty questions were assigned dichotomous closed responses. Similar response options had to be assigned to most of the questions due to the field constraints of requiring brevity and simplicity. More than for the planning questions, the trade off between receiving a superficial broad image and seeking a richer understanding was apparent in the questions on social networks. However it was expected that this would have to be addressed through other means since previous researchers had also found typical questionnaire  93 methods to be inadequate in the Caribbean (Barrow 1983). 5.2.1.4  Execution and analysis The numbers of persons in the various positions who were actually interviewed differed  from those in the quotas shown in Table 5.2 due to the levels of activity observed. First, a larger number of boats than usual did not fish in the early part of the season, mainly due to ongoing repairs. Therefore, the actively fishing fleet was smaller than expected, and consequently the numbers of non-fishing boat owners sought was reduced. Second, large flyingflsh catches early in the season coupled with high unemployment in alternative sectors brought out a larger number of fish vendors than expected, so a number larger than the original quota was interviewed. Third, the latter reason affected the numbers of fishers, and competition to go to sea was fierce, so the numbers of fishers interviewed was more than in the quota. The interviews were conducted between 3 January and 30 March 1994. Each lasted on average about 14 minutes, but ranged between 7 and 64 minutes. Of the 203 respondents, 180 were male and 23 female. The refusal rate was low and mainly due to busy fishers who said that they did not have the time, and older female vendors who were reluctant to talk under any circumstances. Younger females and all male vendors were more cooperative than the latter, but less so than fishers, owners or processors. Some initial hesitancy was encountered when people thought that this was yet another local university or government survey in which they had already participated, but had heard nothing of afterwards. However there was little non-response once the interview commenced, and only twice did an interview have to be completed at another time. Interference or distraction from bystanders was negligible. Throughout the research, despite explaining my student status, people in the fishing industry identified me primarily as a fishery officer (my previous post). Key informants in the industry thought that my association with the Fisheries Division prompted fisherfolk to cooperate and be  94 honest with me more than they would with someone not so affiliated. I have no evidence of this, and cannot say what other biases were introduced because of my perceived position. Post-survey data processing followed standard procedures (Borque and Clarke 1992). The completed questionnaires were checked and edited daily as a quality control measure. Data entry and coding of open-ended responses commenced mid-way through the survey fieldwork, and preliminary data analysis took place at the end of QFISH in order to use the results to design the subsequent research methods. The microcomputer version of the Statistical Package for the Social Sciences (SPSSIPC+) was used. Although some elements of randomization were used so as to improve the applicability of the study results, the fact that respondents were not selected entirely on a probability basis means that the results will be particular to the samples, not necessarily representative of the industry. For this reason the data were subjected only to simple descriptive statistical analysis. Since the results provide information where none existed, and may reveal avenues for further detailed investigation by indicating features useful for to fishery planners to consider, advising the consumer to beware of assuming representativeness is preferable to not collecting or analyzing the data. 5.2.2  Social network measurement The main purpose of this small survey, code-named SOCNET, was to obtain detailed  information on social networks and relations in the fishing industry not attainable through QFISH. The issues on which SOCNET focused were determined from the partial analysis of QFISH data. In addition to the previously cited guidelines on social surveys, literature specific to social network analysis (e.g., Marsden 1990; Scott 1991; Wellman and Baker 1985) was consulted.  95 5.2.2.1  Respondents  A subsample was drawn from among the QFISH respondents on the basis of their willingness to participate in a longer interview. This was assessed in the last question of the survey. The subsample was representative insofar as it was chosen to include a diversity of QFISH responses and all respondent categories. But, since this exercise required much more explanation and effort on the part of respondents, selection favoured respondents that the researcher found to be most articulate and interested in the previous survey. 5.2.2.2  Sampling sites  For several reasons only fisherfolk Bridgetown and Oistins were sampled. First, focus on only two sites relatively close to each other reduced operational constraints. Second, these public markets, because of the number and diversity of users and administration by the state, offer the best opportunity for investigating a rich array of interactions, many of which superficially seemed to be conflictual. Third, since the proportions of male to female vendors were dissimilar at these sites, there was potential for studying gender issues. Fourth, it was at these two major landing sites that both users and the state seemed to be putting most of their efforts into planning improvement. It was here, therefore, that priority appeared greatest for possible application of the research results by planners. 5.2.2.3  Instrument  My requirement was for an instrument that could be administered in about 40 minutes under conditions similar to the previous survey. The instrument needed to cover the following areas: • Composition of informational and instrumental networks by gender, kinship, friendship, linkages and categories of alters; • Information on ties offering assistance through information exchange, marketing arrangements and credit provision;  96 • Indicators of conflict in relations and the types of reciprocity; and • Impressions of what people thought were the important features of social relations in the industry that were necessary to know in order to understand its dynamics. Although network analysis was only used in a limited manner, several methodological issues not encountered in typical surveys had to be addressed. These are reviewed by Marsden (1990), and some of the issues most pertinent to this study are discussed below. The SOCNET instrument, finalized after pretest on three fisherfolk, is in Appendix B. Even though a questionnaire soliciting self-report on social ties through unaided recall is the predominant personal network research method, there are important differences in detail between studies (Marsden 1990). The name generator is of particular conceptual significance. This is a statement used to prompt the respondent into listing the names of their network members according to specific research criteria. My name generator was designed to solicit names of alters involved in routinized ties of an instrumental or informational nature actually used for coping with uncertainty. Actual specific exchanges and routinized ties were chosen to yield improved respondent recall and accuracy. The problem of static bias was addressed later through unstructured interviews which sought to ascertain network dynamics. Name interpreter items addressed both the attributes of the alters and the properties of the ties. However, because of priority and time constraints these excluded common network measurements like frequency of contact, duration of acquaintance and strength of ties. Included due to significance in the fisheries literature review, although known to be methodologically problematic, were measures of conflict and reciprocation. With conflict the problem is one of getting accurate responses to sensitive issues, and with reciprocation there is usually bias towards claims of giving more than is received (Marsden 1990). No boundary was set on networks in terms of the numbers of alters that could be named,  97 network size was allowed to vary across respondents. A list to accommodate twenty entries was chosen on the basis of common research practice, data handling capacity and the pretests. Given research resource constraints it was not feasible to interview named alters although this would have helped considerably to validate ego self-reports, especially about reciprocation. However a question on the perceived amount of interaction between alters was used to obtain an index of linkage in networks. The entire interview was tape recorded in order to capture the vocalized train of thought of the respondent throughout. It was especially useful for capturing the response to the final openended question which in the pretests produced considerable useful information as respondents slipped into a more natural conversational mode. 5.2.2.4  Execution and analysis  The 37 respondents interviewed between 18 June and 15 August 1994, included 26 from the original fishing industry survey. The remaining 11 were fisherfolk not previously surveyed, but required due to the high refusal rate of QFISH respondents in this second round of questioning. However, since care was taken to select additional respondents similar in occupation, age and gender to the original subsample, the final sample appears to be no less representative than the original. Because of their length these interviews were not opportunistic like the first round, but were arranged by appointment. No one openly refused to be interviewed, but three failed attempts to secure an appointment were taken to constitute refusal and that person was removed from the list. The reasons for refusal can only be speculated upon. Secured interviews generally proceeded as intended. In the final sample there were 18 fishers, 10 owners, 7 vendors and 2 processors. Among these 37 were 34 male respondents. It was expected that respondents may have been concerned about the confidentiality of the  98 tape recording, and that reaction to the recorder may have caused bias. However, no respondents objected, some said that confidentiality did not matter, and all appeared to ignore the unobtrusively placed machine once the interview started. The completed instruments were checked and edited daily, and the tapes tested for audibility. The conventional questionnaire data were treated as for QFISH. For the social networks, guidelines for handling tie and net relational data were followed (Scott 1991; Weliman and Baker 1985). It was neither feasible nor necessary to transcribe tapes in their entirety. Only key statements, illustrations and quotations were later indexed and documented. Although the real names of alters were not requested, the consistent use of nicknames that had become familiar to the researcher proved to be an asset in being able to identif,r members common to several networks who appeared to be key people in the fishing industry. Some of these later participated in unstructured interviews. This consistency also allowed clearer identification of the 307 alters to whom the 37 egos were linked by 557 ties composed of 792 strands. Only 3 people exceeded 20 entries on the name generator list and the smallest network comprised 6 alters. Survey of government officers  5.2.3  The main purpose of this survey, code-named QGOVT, was partly to solicit opinions on several of the areas investigated by QFISH, but mainly to investigate attitudes towards planning and the involvement of the fishing industry. The usual literature on social surveys was again consulted (see above). 5.2.3.1  Respondents  All government officers likely to have any influence on, or to participate in, fishery planning were selected for QGOVT. This amounted to a census of 13 people in the Fisheries Division and the administrative section of the IVlinistry of Agriculture, Food and Fisheries.  99 5.2.3.2  Instrument  My requirement was for an instrument that could be administered in about 40 minutes to cover the following areas: • Respondent categorization such as position, gender, age, and education; • Uncertainty and social strategies for coping; • The relationship between fishery management and development planning; • The role of the fishing industry in fisheries planning; and • The availability and potential uses of biological, social, and economic information on the pelagic fishery for planning. The QGOVT interview guide provided at Appendix C was designed without the benefit of a pretest due to the small number of potential respondents, but it was partially based on the results of QFISH. The time limit reflected the longest anticipated period without interruption to be expected in the government offices where the interviews were to be conducted. Office interviews were selected to facilitate sequential batch interviews where possible in order to reduce the possibility of discussion of the interview by prior respondents influencing potential respondents. Based on the experience of Khan (1987) in interviewing Caribbean public servants, a semistructured interview format was followed to allow latitude in responses. The interviews were designed to be tape recorded in order to capture this latitude in response. A few questions were identical to those in the fishing industry survey in order to facilitate direct comparison. Since several of the questions addressed complex topics, closed responses were provided on the interview guide to provide prompts for the common ground that had to be covered. 5.2.3.3  Execution and analysis  Permission from the Permanent Secretary of the Ministry was sought for the interviews in  100 order to assure officers that the time spent with the researcher was approved. Access to the officers was not problematic, and although it was not possible to arrange the interviews in batch sequence at the two offices there was no detectable contamination from previous interviewees. Seven of the 13 interviews were with Fisheries Division staff, and there were four females in the census. The interviews lasted between 31 and 111 minutes, averaging 65 minutes. As found by previous researchers (Lindsay 1978; Khan 1987), the civil servants spoke very freely and several seemed to appreciate the opportunity to discuss their work in detail. Special attention was paid to detecting respondent bias from being previously acquainted with the researcher, but none was obviously discernible. Analysis of QGOVT was similar to that for QFISH in terms of the questionnaire, and to SOCNET in terms of the tape recordings. Other interviews  5.2.4  The instruments described above were essentially structured questionnaires even though flexible use as interview guides was incorporated in the latter two. Less structured were interviews conducted by appointment with persons involved in fishing industry organizations, and the ad hoc unstructured and informal interviews held with key informants and other fisherfolk at fish landing sites. 5.2.4.1  Semi-structured  During the earlier research several sources pointed to formal organization as the route many people in the fishing industry favoured to address their perceived uncertainties and problems. Previous experience with fonnal organizations also seemed to have shaped their attitudes towards working with each other and government. Based on these findings, interview guides were the main tools chosen to investigate the formation and operation of early fisherfolk and government organizations through questioning the key people involved. Since some of these  101 events occurred up to 30 years ago, making the accuracy of respondent recall an issue, they were each important for validating the other and the scant documentary evidence available. The key informants were identified through a reputational or snowball sampling process (Bailey 1982), and all 10 persons selected consented to be interviewed. The focus was on fishing cooperatives, a trade union, the Cooperatives Division and the Barbados Marketing Corporation. The interview guides used varied with respondents (all male), but covered topics such as how and why the organization was formed, who were its members, what were its objectives, what did it achieve and how, what were the interactions with sections of the industry and government, and what problems were experienced and why. The interviews, conducted near the conclusion of the fieldwork, were all tape recorded and lasted between 30 and 200 minutes. The audio tapes were analysed as previously described. Special care was taken to detect sources of error  such  as  impression management,  topic  avoidance,  deliberate  distortions  and  misunderstanding (McCracken 1988). While some impression management was noticed in the form of self-aggrandizement there were few apparent serious sources of error. 5.2.4.2  Unstructured and informal  Early in the research a few fisherfolk were selected as key informants, or consultants, on the basis of their long experience in the fishing industry, their ability to analyse and articulate concerns, and their interest in keeping track of the research. In order to minimize potential bias, these people were not explicitly informed of their special status, but were frequently consulted in the form of unstructured interviews to assist in the interpretation of observations and results. Throughout the fieldwork informal interviews were held with a wide variety of fisherfolk often during periods of participant observation as described below. Both types of interview lasted from minutes to hours, but none was tape recorded. The  102  main points were entered into field notes as soon after the interviews as practicable. Due to the natural atmosphere and unlimited flexibility, compared to the previously described more structured and quantitative ones, these interviews provided a far richer source of qualitative material essential for understanding the meanings of communication, actions and events from the perspective of fisherfolk. Participant observation  5.2.5  Although participant observation was not intended to be a primary research method, like the unstructured and informal interviews it turned out to be important. In order to capitalize on the validity potential of this method, relevant guidelines (Jorgensen 1989) were followed, and the caution that the researcher be aware of his social location was kept in mind. 5.2.5.1  Fish landing sites My perceived social location as a fishery officer precluded completely blending in, but  according to key informants, participating in the normal activities and conversations at the two main fish landing sites facilitated my acceptance and elicited greater cooperation than would otherwise have been received. My focus on Bridgetown and Oistins was for the reasons listed previously (see SOCNET sampling sites) plus the fact that it was only here that there was sufficient activity in the form of people of all types just hanging around, observing and talking to each other for the researcher to not be overly conspicuous. Between interviews and at other times I closely observed interactions and transactions between fisherfolk, but particularly among fishers and between fishers and vendors. Observational evidence of the types of relations described in the QFISH and SOCNET surveys were sought and recorded. I also took note of fish landings and prices in the field notes compiled afterwards.  1 -•1  lv.,  5.2.5.2  Organization formation  A fortuitous unforeseen event, described in detail in Chapter Eight, was the emergence of a fisherfolk organization formed initially as a response to a specific perception of uncertainty, and later conceived generally as a response to several perceived existing and potential sources of uncertainty. Due to prior acceptance by most fisherfolk I was able to observe and participate in the formation process at a time when government officials and some fisherfolk (specifically processors) were excluded. This research was conducted between 12 April and 27 August 1994. It facilitated the collection of additional firsthand data on perceptions of uncertainty, the notions behind the use of the formal organization as a coping strategy, the effect and dynamics of social networks within and outside the organization, and the problems and potential of interaction with the state. For the study, this event was an opportunistic natural experiment during which meeting minutes and supplementary field notes were recorded. 5.2.6  Document study  Document sources included academic and government institutions, CFRAIVIP and the popular press. The acquisition of natural and social scientific literature on phenomena perceived to be uncertain, and pertinent Caribbean sociology, was straightforward from all sources, but not abundant. The paucity of relevant documentation was encountered repeatedly. The difficulty, mentioned by other Caribbean researchers (Lindsay 1978; Khan 1987), with access to government documents such as files, statistics, development plans, political manifestos, annual reports and the like was not encountered. Possible reasons include my status as a government official, my knowledge of what material did or was likely to exist, and the remoteness of any of the information requested being considered sensitive. In subjecting the official documents to content analysis, the possibilities of errors, omissions and bias were noted (Bailey 1982).  104 5.2.7  Discussion The research design incorporated several methods to attempt to answer the questions.  Multimethod research or triangulation tries to compensate for the weaknesses of one method with the strengths of another (Brewer and Hunter 1989). An example was the use of unstructured interviews and participant observation to provide the qualitative depth missing in the rapid questionnaire survey. Since theory-construction or hypothesis-testing are not the prime objectives of this study, and the situation in Barbados is expected to be dynamic, less emphasis was placed on reliability and external validity than accuracy and internal validity. Qualitative methods were used to try to ensure that issues were understood and the research concepts valid. But without the data from the more quantitative methods it would have been difficult to determine which issues were the most important on an industry-wide basis. Such reinforcement of methods by sequence and type may be particularly useful in conditions similar to those in Barbados about which little is known and data are scattered or scarce. Lessons were also learnt from the execution of the methods. For example, fisherfolk are willing to participate in social surveys partly because they feel that the results may be used to improve their circumstances. However, the patience of some appears to be wearing thin due to the increased frequency of questionnaire surveys coupled with the lack of feedback from them (perhaps symptomatic of a deeper communications problem). The academic and government institutions particularly guilty of this are likely to soon create an environment hostile to integrating social science data into fishery planning through the survey method. Yet, so far this reaction is entirely against questionnaires which, in any case, may not be the most appropriate way to collect social science information or identify problems or solutions. This is especially so in the complex and unfamiliar territory of fisheries management planning. When confronted with a questionnaire, fisherfolk are adept at supplying the minimum  105  information required, and typically do not seek clarification or explanation when confusing or unfamiliar terms arise. Because of this, and since pockets of illiteracy eliminate other than personal interviewing anyway, it could be feasible to use unstructured approaches (perhaps with a tape recorder) in order to cover similar ground more effectively, and with more meaningful results.  106 6. FISHERFOLK AND UNCERTAINTY IN THE PELAGIC FISHERY  In this chapter I address the first research question mentioned in Chapter 2, that is: what uncertainties are perceived to characterize the Barbados pelagic fishery. I begin by describing the salient characteristics of people involved in the fishing industry. Next, I turn to perceptions of uncertainty reported by fisherfolk and state officials in the QFISH, QGOVT and other interviews. Then, scientific data and official records related to the factors perceived to be most uncertain are considered. This is done in order to compare these technical measurements with the perceptions reported by fisherfolk. Before concluding, I describe a simple exercise that examined the potential for reducing uncertainty through information exchange between the state and fishers. Tables of the results discussed are provided in Appendix D if not in the chapter. 6.1 SIJRVEY RESPONDENTS  I begin with a profile of the fisherfolk and state officials who were respondents in the QFISH and QGOVT surveys respectively. The QFISH survey was conducted primarily to quickly provide some basic background data on the industry. The QGOVT survey was more detailed and focused mainly on fishery planning. 6.1.1 Occupational categories  The assignment of fisherfolk to the occupational categories of fishers, owners, vendors and processors was a critical first step in the analysis of the QFISH survey data since the manner in which respondents are categorized has implications for interpretation. In all of the above categories some respondents fill more than one occupational niche in the industry. In addition to theoretical considerations, the finding from pretest questions that people invariably identified more with one aspect of their work if they occupied multiple niches guided selection of  107  occupational categories as explained below. The category ‘fishers’ for example, comprises captains and crew who fish on different types of boats (Table 6.1). Occupational categorization by position and boat type was not used in analyses since fishers are quite mobile within the fleet. Crew readily move between boat types, and dayboat captains may serve on iceboats as crewmen for the anticipated higher income. In the offseason mobility is even higher as fishing becomes more of an opportunistic activity and switching between fisheries is frequent. For these reasons, and the prevalence of good relations with low conflict among boat types (examined later), a single category was used for fishers. Table 6.1 Distribution of fishers by boat type Boat type fished on Davboat Iceboat Iceboat/Longliner Longliner  Captain  Crew’  (%)  (%)  66 30 1 3  69 23 4 4  Sampie (N Note: Unless otherwise indicated, in all tables the column percentages sum in each column to 100% which represents the sample size for the column category.  Boat ownership sub-divided all occupational categories (Table 6.2). Although this research was not undertaken from a political economy perspective, evidence that ownership of capital influences socioeconomic relations has been well established in the literature (e.g., Barrett and Apostle 1989). In the pretest it was found that fishers who owned any type of boat perceived themselves primarily as fishers, not boat owners. Most of the dayboats owned by the fishers were older or smaller vessels with low capital value, The self-perception of fishers who owned iceboats, egalitarian relationships between fishing owners and crew, and the fact that boat ownership did not necessarily imply high net income (also examined later) argued against putting iceboat fisher-owners into a separate category.  108  Table 6.2 Boat ownership among respondents Types of boat owned None Moses Dayboat Iceboat Iceboat/Longliner Longliner Dayboat & Iceboat Dayboat, Iceboat & Longliner Dayboat & Longliner  Captain  Crew  Owner only  Vendor  Processor  (%)  (%)  (%)  (%)  (%)  52 1 36 9 1  90 4 2 2 2  97  40  1  48  65 20 3 5 3 2 2  60 3  32  Vendors predominantly had one occupation and none of the processors were the sole owners of the few longliners in which they had invested. Given the small number of processors in their census, further sub-division of this category was not undertaken. The term ‘owner’ was hence reserved to mean boat owners who were not otherwise involved in the fishery, that is, who were not also fishers, vendors or processors. Breakdown of owners by boat type was not necessary since their experiences and outlooks were relatively similar despite the differences in their levels of estimated capital investment. The latter were too inexact to serve as a useful basis for categorization. In this manner the four main occupational categories used in this study were selected. Preliminary analyses also revealed that little additional information from the QFISH survey data would be gained by disaggregation of occupational categories. Throughout this study, since respondent selection was not random, the survey results are not necessarily representative of fisherfolk in general. However, in reporting and interpreting results, attention is paid mainly to the major trends and features of findings that may have the widest relevance. 6.1.2 Respondent profile  The QFISH sample reflects the preponderance of males in the pelagic fishing industry.  109 Nationally, across all industries, women comprise 60% of the labour force (Government of Barbados 1993). Compared to the national statistic, males dominate all fishing industry occupations except for fish vending where the female participation rate is 62% (Table 6.3 a). But  the latter differs by site, female vendors being more in evidence, for example, at Oistins (8 5%) than at Bridgetown (5 0%). The marked sexual division of labour reported here is common to fisheries around the world (Acheson 1981), and differs little from earlier accounts of the Barbados fishery (e.g., Stoffle 1969). The female government officers were in relatively senior positions. Those in the Fisheries Division had begun to study fishing industry gender issues such as the savings practices reported on later. There is little difference across industry occupations in the mean age of fisherfolk, although boat owners tend to be older (Table 6.3b). This probably occurs because young people have less opportunity to acquire the capital necessary to buy a boat. Also, as mentioned in Chapter 4 and discussed later, owning a boat may not always be as profitable an undertaking as implied by the estimates in Table 4.5. Hence young people, especially those familiar with the industry, may see boat ownership as a burden to be avoided. The relative scarcity of young boat owners may have implications for innovation and related factors that are not prime considerations in this study. The mean period of involvement in the industry was also similar among fisherfolk sampled, except for the more recent development of fish processing which commenced in the 1970’s (Table 6.3c). Before then one mainly had different scales of fish vending. Many of the owners who have less than average experience in the industry own iceboats, a section of the fleet which expanded rapidly in the 1980’s. Within government, the staff of the Fisheries Division are on average younger and have been slightly longer in their present jobs than those in the Ministry, but there are few officials with long fisheries service. This latter finding may be important in terms of institutional memory  110 and attitudes towards the industry. Officials’ perceptions of state intervention and the industry may lack the historical perspective of experience. Although Ministry staff do not deal frequently with fisherfolk in their normal line of duty, they all said that they were regular buyers of fish at markets. Since vendors handle most fish sales, the Ministry officials’ impressions of the other fisherfolk, especially fishers, may not be based on direct interaction, but on social stereotyping. Although experiences and impressions may either be positive or negative, the social image of fisherfolk held by the public (discussed in Chapter Seven) is predominantly negative. Hence this may be the one more likely to be received by Ministry staff Conversely, fisherfolk may hold stereotypical images of government officials as self-serving and oppressive of the lower classes (Khan 1987), and unable, due to lack of experience, to communicate on fishery matters. In the industry, owners and processors tend to have more formal education than fishers and vendors, but fisherfolk typically have less formal education than government officers (Table 6.3d). While the }Vlinistry has proportionally more officers with tertiary education, none have the technical fisheries training received by over half of the Fisheries Division staff In random samples of the general population taken in 1978 and 1984, the proportion of people with primary education decreased from 58 to 50%, while secondary increased from 40 to 44% and tertiary from 3 to 4% (Dann 1984). If my sampling was random by design, and the trend towards more formal education continued since 1984, one could conclude that fishers and vendors have on average less formal education than the typical population, and that all other respondent categories in Table 6.3d have more. The distribution of education amongst owners reflects recent entrants with higher education mixed with older, less educated owners. Education is an important background variable for several reasons. Most obviously, persons with less formal education have a limited range of economic opportunities since jobs requiring  111 paper qualifications are not available employment options. However, Dann (1984) found that although education was popularly perceived as being important to Barbadians, people randomly surveyed ranked it low in their hierarchy of “important things in life” (p. 100). Education was perceived to have little impact on their daily lives, but was highly correlated with social class and seen as an avenue to individual social mobility. If these attitudes and perceptions hold true today, it is possible that fisherfolk could perceive the higher formal education of the state officials, not as an index of competence and expertise, but mainly as a social barrier. Some fishers, vendors and older owners were sensitive to queries about their level of education. There was a strong impression that authority was not accepted automatically by fisherfolk on the basis of title or education alone. Informants said that only if highly educated individuals were perceived as practical and “down to earth” were they likely to be taken seriously by fisherfolk. Wilson (1973) reached similar conclusions for other Caribbean coastal communities. He found that competent persons who were not socially accepted into the networks of lower class clients were ineffective as change agents. While being “down to earth” is a primarily a matter of social and communication skills, what is perceived to be practical depends on education and experience. These findings could be especially pertinent to interactions between state officials and fisherfolk given their different backgrounds. Tables 6.3(a) to (d) Profiles of QFISH and QGOVT respondents  (a) Gender Gender Male  Female Sample (N)  Fisher  Owner  Vendor  ()  °4)  Processor Fish. Div. Mm. Agric.  ().............  99 1  95 5  38 62  100  57 43  83 17  126  40  32  5  7  6  112 (b)Age Age (years) under 24 25to34 35to44 45to54 55to64 65to74 over 75  Fisher (%) 7 18 28 27 12 8  Owner (%)  Vendor Processor Fish. Div. Mi Agric. (%) (%) (%) (%)  3 18 34 24 16 5  10 39 29 16 7  20 60  57 29  50  20  14  33 17  Mean  44  54  46  42  36  46  126  38  32  5  7  6  Vendor  Processor  (%)  (%)  26 22 26 13 13  40 40 20  Fish. Div. (%) 100  22  23  14  12640  .1  Sample (N)  (c)  Period of involvement Period of involvement in fishing industry (years) under 10 llto2O 21to30 31to40 41to50 over 50 Mean  Fisher Owner (%)(%) 37 25 27 17 25 15 14 15 6 8 2 8 21  Sample(N)  Mm. Agric.  (%) 100  4  2  (d) Formal education Education Primary  Some secondary Full secondary Tertiary Sample (N)  Fisher  Owner  Vendor  (%)  (%)  (%)  61 18 20 1  50 10 28 12  66 22 12  126  40  32  Processor Fish. Div. Mi  Agric.  (%)  (%)  (%  20 60 20  14 43 43  17 83  5  7  6  6.2 PERCEPTIONS OF UNCERTAINTY  I focus mainly on the uncertainties that people perceive and can express in some quantifiable form. These perceptions are compared with reported measurements. Often, for example in the  1  I I.  case of environmental and ecological variables, this is a comparison of ordinary and scientific knowledge. The variables of prime interest here are perceptions of uncertainty across the occupational categories discussed in the previous section. 6.2.1 General uncertainty One of the QFISH questions asked respondents to rank the extent to which they agreed or disagreed that the fishery was generally unpredictable. The modal response among government officers and processors was “agree” (57% to 67%), and among owners, fishers and vendors it was “agree very much” (41% to 70%) (Table 6.4). This suggests that uncertainty is generally perceived to characterize the Barbados pelagic fishery. Table 6.4 General perception of uncertainty by occupation The pelagic fishery is geral1y unpredictable Agree very much Agree Not sure Disagree  Sample (N)  Fisher () 62 24 2 12 126  Owner .°<) 70 18  Vendor  12  41 34 3 22  40  32  Processor Fish. Div. Mi Agric. (%) ( () 60 40  14 57 14 14  33 67  5  7  6  Disagreement was highest (22%) among the vendors. Some people, mainly fishers and vendors, laid off from other industries disagreed since they found self-employment in the fishing industry to be more reliable for providing income. These were invariably recent entrants. In the Fisheries Division two respondents argued that the pelagic fishery was not unpredictable, but seemed complex, and chaotic. 6.2.2 Specific uncertainties The uncertainties perceived by all categories of fisherfolk are examined first, followed by more detailed analysis of the harvest and post harvest sectors. Finally, the views of state officials are noted.  114 6.2.2.1 Alifisherfolk Over 95% of fishing industry respondents perceived total catch, proportions of species in catches and their earnings (daily and seasonal) to be unpredictable (Figure 6.1). About 90% found the kinds of fish expected in the catch and fish prices unpredictable, but around 75% thought that season duration and periods of peak abundance for particular species were predictable. The pelagic fishing season was perceived to run from November to July, although longliners attempted to fish year round. Periods of peak abundance were identified as May for flyingfish, and February to March for dolphin and kingfish, but tuna and billfish could be abundant in any month. Agreement on these periods among respondents was high.  Figure 6. 1 Uncertainties perceived by all fisherfolk  Are you usually able to predict:  N=203  a  thy or trip earnings  j=’  fish prices season end  peak abunthnce  Qno  a_________  season earnings  sesason start  Dyes  ?‘t. I Ie2IIIIIIIl2.%.p.II ‘.2.2.2.2.211 ‘22222.22  ’222%1t 2  2.?.c.% 12.222:62 ‘c22€€62  62P2t?t2 ‘t22  species proportions  =  species in catch  •2  221222 P222622  a  ==;;;  ----  $  total catch per trip 0%  10%  20%  30°/d  40%  50%  60%  70%  30%  90%  I00°’s  Percentage response  During informal and unstructured interviews it appeared that the relatively new practice of longlining was redefining the fishing season and perceptions of fisheries ecology. Iceboats had  115  done this a few years earlier by increasing access to new fishing areas. The impression was that pelagics could be available for more of the year, but further afield, than previously thought. Just as iceboats had extended the fishing season, so too were longliners, and more iceboats were being outfitted to switch to longlining in the offseason. 6.2.2.2 Harvest sector  Fishers, when asked about uncertainty in their routine operations, reported equipment failure such as gear loss and mechanical breakdown to be unpredictable (Figure 6.2). Some fishers said that environmental factors such as rough seas and current direction were predictable (5 and 13% respectively). Slightly more fishers said they could predict seasonal fish migration  and good fishing locations (17 and 22% respectively). Fishers’ observations on search methods and currents were more detailed than those on fish migration, but most were either unwilling or unable to systematically organize and link their observations (in a way that I could understand).  Figure 6.2 Uncertainties perceived b fishers  Are you usually able to predict:  ayes  no  N=126  rough seas gear loss breakdown fish migration current direction •:  fishing location 0%  10%  20%  3000  40%  50°o  60%  70%  80%  90%  100%  Percentage respoisse  The search methods included paying attention to moon phase, water colour, drifting objects  116 and seabirds since fish catches were thought to vary with these, often inter-related, factors. They did not explain their observations in terms of the major currents and oceanographic processes in the region. Many fishers added, however, that their perceptions of uncertainty depended not only on their skills and knowledge (human capital). They also depended on their ability to be in touch with others who could supply reliable information on recent catches (social capital) as described in the next chapter. The hunt for fish by North American fishing boats is heavily dependent on sophisticated technology. Cost, size and boat design limit dayboats to being equipped usually only with radios, but most iceboats carry more wheelhouse electronics. Yet the potential of available onboard instrumentation such as fish finders, water temperature gauges and global positioning systems (GPS) to reduce uncertainty is not fully exploited. This appears to be due to lack of education, training, capital, and a preference to rely on intuition and observation. In many cases electronic instrumentation is not installed by owners since it will not be used by fishers. Some fishers described the admiration for their ability to successfully “fish blind” which they received from American longline skippers who set gear only when their instrumentation reduced uncertainty considerably. Comparison of the results of blind and instrument-guided fishing effort was not possible. The complex topics of fisheries conservation, ordinary knowledge, and attitudes towards scientific knowledge relating to uncertainty were tackled mainly through unstructured and informal interviews. Of all the pelagics, only flyingfish elicited concerns about conservation, and this was only from a minority of fishers and boat owners. They were aware that the fishery targeted spawning aggregations and were concerned about this. They wondered if retrieving instead of releasing the “screelers” (tethered fish aggregating devices), when they were laden with recently spawned eggs, could reduce future abundance. Given that the use of new screelers  117 on each trip would impose an additional input cost with unknown returns to the fishing enterprise, they were uncertain about its feasibility as a conservation measure. Fishers revealed a wide range of knowledge on all the pelagics based on frequent observations at sea and examination of the fish ashore. Observations included fish abundance and distribution, behaviour, stomach contents, external features such as body shape and markings, and relationships between catch and oceanographic features. They had few explanations for their observations and said that scientific knowledge could be useful to them. But they also doubted that Caribbean marine scientists knew as much about “out there” as did fishers, since they were seldom seen at sea. No fishers interviewed were aware of regional or local research findings such from the Eastern Caribbean Flyingfish Project, although most were aware of its tagging program and had some idea of its purpose. Since academic and government researchers have used local fishers and fishing boats in studies, some had participated in fisheries research. But the interest to be told about the findings was not great among any of the respondents. Few fishing and non-fishing boat owners (only 11%), reported uncertainty about where the catch would be landed at the end of a trip (Figure 6.3). Yet, the landing site often changed during the season, especially for the east coast boats, when seas became rough at their home ports. Owners who were uncertain about their landing site explained that sometimes their boats either went to wherever the price was best, or were not sufficiently powerful to travel against the current at times so landed at the closest port. Most owners were certain (67%) about finding regular crew, although uncertainty about crewing was highest among dayboat owners. The latter experience a high rate of crew turnover due to the relatively unskilled nature of the work. Only 38% of owners reported being certain about their ex-vessel fish buyers, and more were uncertain  about short term (trip) profits than longer term (season) profits (83% versus 68%).  118  Figure 6.3 Uncertainties perceived by fishing and non-fishing owners  Are yon usually able to predict:  Dyes D  N=82  landing site regular crew  EE’  regular buyer  cc  .  vv  supplies availability  a  re E  t-  3  — — — a — — F za — F F  trip profit  ,.  :  .  season profit 0%  10%  20%  30%  40%  50%  60%  70%  aF a  80°/d  90%  100%  Percentage response  Most owners were also fairly certain (73%) about fishing supply availability, but iceboat owners reported relatively high levels of uncertainty about getting sufficient ice promptly. This problem pertained to those who relied mainly on the government’s cheaper supply at Bridgetown. What boat owners and fishers argued was that, unless fish abundance and location were fairly certain, it made more sense to wait for the cheap ice (lower operating cost) than to risk a poor trip with expensive ice. The larger the boat, the greater was the incentive to reduce uncertainty by acquiring information before going fishing. It appeared that sets of iceboats went to sea and returned at the same time which increased their ice, price and hence income uncertainty. 6.2.2.3 Postharvest sector In the postharvest sector, when asked about operational uncertainties, vendors and processors were less certain about making a seasonal profit than boat owners (43% of vendors compared to 68% of owners) (Figure 6.4). Less than half of the vendors and processors said  119 they could predict good sales days, fish prices to customers and the amount of fish spoilage (42%, 28% and 31% respectively). However, 80% were certain about which boats would expect them to buy their fish, and 54% knew which boats would not sell their catches to them. Vendors were often heard to refer to boats as “theirs” or someone else’s. This referred not to ownership, but the shared knowledge of who would be offered the catch first because of several types of informal ties with the fisher or owner. Figure 6.4 Uncertainties perceived by vendors and processors  Are you usually able to predict:  LI yes  N37  LIno  =  season profit t  amount 0 f spoi.1 age goodsales day resale prices  r z: . _,::..e,l. ,:c.r;.rf,;  Z  boats expect to sell boats will not sell  t  == = = = =, = = = i  .  9:  0%  10%  20%  30%  40%  50%  60%  70%  80%  90%  100%  Percentage response  Since the QFISH pretest showed that state intervention was a complex and provocative issue, this was explored through less structured interviews. Many fisherfolk were unaware of the frmnctions and actions of the state agencies responsible for the fishing industry and the recent, comprehensive Fisheries Act (1993) which governed the fishery. Since the state was mainly perceived to neglect rather than actively intervene into the fishing industry, there was the perception among fishers, vendors and some owners that it did not generate much uncertainty. However, what uncertainty it did generate was commonly perceived to be in the form of constantly changing policies and procedures. Processors had to deal more with the state because of  the nature of their business. They all felt that a high level of government intervention and  120  uncertainty affected them. 6.2.2.4 State officials Detailed perceptions of uncertainty in the fishing industry were not solicited from government officials, but they were asked for general impressions. Most in the Ministry thought that ecological factors and marketing practices contributed very much to uncertainty (both 66%) (Table 6.5). Most Fisheries Division respondents thought these contributed relatively less (57% and 43% respectively). In the Ministry, fluctuation in catch was seen “as just luck from year to year” (QGOVT Interview No. 11). Conversely, officials in the Fisheries Division thought that fishing methods and state intervention contributed more to uncertainty (43% “very much” and “much” respectively) than did people in the Ministry (33% “much” and 66% “fair bit” respectively). Table 6.5 Government officers’ perceptions of contributors to uncertainty Fisheries ecology  Fishing methods  Extent of  Fish. Div.  Mm. Agric.  Fish. Div.  contribution  (%)  (%)  (%)  Little Fair bit Much Very much Don’t know  29 57 14  Sample (N)  7  17  29  17 66  29 43  Mi  Agric.  (%) 17 33 33 17  6  7  6  Table 6.5 continued Fish marketing Extent of Fish. Div. Mm. Agric. contribution (%) (%) Nothing 29 Little 14 Fair bit 14 17 Much 43 66 Very much 17 Sample (N)  7  6  State intervention Fish. Div. Mm. Agric.  (%) 14 43 43  7  (%) 17 17 66  6  Regarding state intervention, the lack of clear and consistent policy, and the gap between  121  plans and practice, were cited as sources of uncertainty. Referring to the development plan, one official said that “in writing there is a policy, but I’m not too sure this policy is being implemented” (QGOVT 10). For example, the harvest sector relies heavily on subsidy through concessions to import goods free of duty (e.g., boat equipment and fishing gear). But another official remarked that the concessions “become available and then they disappear, and then you are not sure whether they are available or they aren’t” (QGOVT 11). 6.2.3 Major uncertainties  The aspects of the fishing industry perceived by fisherfolk and state officials to be most uncertain are shown in Table 6.6. Fish prices and catches were reported most frequently as the most uncertain aspects of the industry. The majority of fishers and vendors thought prices were most uncertain, but owners, processors and the Fisheries Division put catches first. Ministry officials were split between the two. Boat owners, vendors and processors were the only ones who found customer demand to be a significant uncertainty. Many reported that this was a recent, and presumably temporary, uncertainty caused only by the general economic conditions in the country. Demand uncertainty, rather than supply, they said had more influence over ex vessel prices, causing them to lower the prices offered even when fish was scarce. Table 6. 6 Aspects of the fishing industry perceived to be most uncertain Most unpredictable  Fisher  1 i ndustry Fish prices Fish catches Customer demand  (%)  Ice availability Weather/sea state Everything  Safety at sea  State intervention Other Don’t know Sample (N)  35 26 8 8 6 4 1 2 10 125  Owner Vendor Processor Fish. Div. Mm. Agric. (/) ( 30 28 43 33 35 13 60 57 33 10 19 20 5 6 5 2 3  9  17 17 20  10  3 22  40  32  5  7  6  122 The prominence of ice availability as an uncertainty was reported mainly by iceboat owners and fishers. Its absence from the concerns of the government respondents is notable since government is the main source of ice. Weather, sea state and ocean currents were prime concerns only to fishers, but they were joined by owners and a Ministry official in concern over safety at sea. State intervention ranked low as a maj or uncertainty and did not figure at all in the responses from vendors and government. Several people said that “everything” about the fishing industry was uncertain, and an unexpectedly high proportion did not know what aspect to say was most uncertain. In a sense these two responses were similar since people frequently said that so many things about the industry were highly variable that they did not know which among them was the most unpredictable. 6.3 OFFICIAL AND SCIENTIFIC MEASUREMENTS OF UNCERTAINTY  The data in this section are derived mainly from official and scientific statistics and documents available to fishery planners. The resulting information is compared to the results from the interviews, with emphasis on the major uncertainties, to determine where they are similar and where they diverge. 6.3.1 Fish catches At the primary landing sites Markets Division staff compile daily records of fish prices and landings (synonymous with catch since discards are negligible). The comments in Chapter Five about the errors and omissions in official statistics should be borne in mind. Observation and document study suggested that only Oistins fish market provided reasonably accurate and complete time series useful for examining variability in fish catches and prices. Although the Oistins data are not necessarily representative, they serve to illustrate features believed not to be unique to that site.  123  Uncertainty about fish catch is multidimensional. Measurements of several of the factors that contribute to fish catch uncertainty, which were identified in the interviews, are described here. 6.3.1.1 Seasonality Although the landings time series goes back 30 years or more, the years 1981 to the present were chosen for analysis. They encompass the recent modernization period of iceboat and longliner fleet expansion easily recalled by respondents, and so provide the best comparison between the perceptions of fishery workers and official records. Figure 6.5 illustrates interannual variability in the Oistins landings of the major pelagic species during this period. Seasonality is obvious even when all species are combined. Since the fishery is multispecies, it is this image that determines the perception of overall season duration and interannual variability. Although there is variation within seasons, from year to year the November start and July end is quite consistent. Over the 14 year period there is no clear trend that total pelagic landings are either increasing or decreasing. Examination of the individual species components (provided in Appendix D) leads to the same conclusions.  124 Figure 6.5 Oistin’s total pelagic fish catches, 1981 to 1984  Multispecies pelagic fishery  Catch (metric tons) 160 140 120 100 80 60 40 20 0 81  81  82 82 83 83 84 85 85 86 86 87 88 88 89 89 90 Year  90  99  92  92 93 93 94  The within-season monthly variation in landings of each species is seen in Figure 6.6(a) to (e). Landings for each month in the time series are plotted with the means of these values in order to illustrate details of the seasonal variation that respondents may recall over a period of several years. Within the season, and particularly around the seasonal peak(s) of abundance, the monthly landings vary substantially from year to year. This is especially so for the two major species, flyingfish and dolphin. Figure 6.6(a) to (e) Seasonality of major pelagic species at Oistins, 1981 to 1994 (a) Flyingfish  Catch (metric tons) 120  —•—--  +  +  100 80  Mean  +  +  +  *  60  +  40 20 0  Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 14onth  125 (b) Dolphin  Catch (metric tons) 90 80 70 60 50 40 30 20 10 0  —•—  +  Mean  +  IIIo  Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Month  Catch (metric tons)  (c) Kingfish —4——Mean  25 + 20 15 10 + )  0  Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Month  Catch (metric tons)  (il) Tuna —•——  7  Mean  ÷  6 5 4;  +  3. 2  +  +  zE  +  +  Hit  Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Month  126 (e) Bilifish  Catch Onetric tons) 4 3.5 3 2.5 2 1.5  Mean + + +  + +  + +  +  +  +  0.5 0  Jul Jan Feb Mar Apr Aug Sep Oct Nov Dec May Jun Month  Using a longer time series from Oistins and more sophisticated analysis, Mahon et al. (1990) reported the seasonality described in Table 6.7 which agrees fairly closely with the data plotted above and the reports of the fisherfolk. The agreement between perceptions and measurement of seasonality is not surprising since they are based on the same data set—the commercial landings. A more informative picture of seasonality could be based on test fishing independent of the commercial fleet. Table 6.7 Seasonality in the pelagic fishery Catch Flyingfish Dolphin Kingfish Tuna Bilifish Source:  Seasonality based on average pattern Unimodal May Unimodal Feb-Apr Slightly bimodal Jan-Jun Bimodal Dec and Jun Slightly bimodal Dec and Apr Mahon et al. 1990  at Oistins  Seasonality based on individual years Frequently bimodal Usually unimodal Frequently bi- or trimodal Usually bi- occasionally trimodal Spiky  Independent of commercial catches, Khokiattiwong (1988) verified flyingfish seasonality near Barbados, finding few or no fish of catchable size between August and November. He also confirmed the unpredictably patchy distribution of flyingfish schools reported by fishers. For the large pelagics, the assumption common in the literature is that availability to the commercial fleet is a reasonable index of abundance, and hence seasonality, within the areas fished (Mahon  127  1987). Independent research on these species is limited by the costs involved. 6.3.1.2 Search parameters  In terms of what fishers search for to find fish, during the early summer months of fieldwork fishers reported high catches of dolphin in green water passing southeast of the island. Several boats were tracking this patch since it increased the certainty of good catches. Reports from fishers and others of higher catch rates associated with green water lenses from the distant Amazon and Orinoco rivers are common (Brown 1942; Muller-Karger 1990; Gomes et at. 1994). Khokiattiwong (1988), however, found no catchlcolour correlation in his research on flyingfish. The occurrence and trajectory of these river water lenses seem to be unpredictable. This is because of the complicated relationships between the volume and timing of the Amazon and Orinoco river discharges which produce them, and the poorly understood ocean circulation patterns which transport them. Thus, finding a rich patch is likely to be unpredictable, but keeping track of it once found is not a difficult proposition if observations on position are made frequently and shared among fishers between trips. Gomes et at. (1994) report that the occurrence of drifting objects, and perhaps the tendency of fish to associate with them, is greatest in water which is green or brown in colour. They confirm that drifting natural or anthropogenic objects, which occur singly or along fronts called “riffles”, are sought by fishers because they have been proven to attract pelagics and increase their catchability. Flocks of feeding seabirds often provide visual clues about the location of these objects from a distance. These authors find that the formation of riffles is most frequent during the January to March period of higher current speeds. But no clear pattern to the spatial and temporal occurrence of drifting objects has been described. Early in the flyingfish study an increase in flyingfish catch per unit effort (CPUE) around the frill moon and a decrease around the new moon was found. This is consistent with the  128 observations of fishers, but was later reported to be a weak correlation (Oxenford et al. 1993). So in this case, as with the other search parameters, it appears that the uncertainty of fishers is reasonable to expect given the large gaps in scientific knowledge. As with seasonality, there is, reasonable agreement between what is perceived by fishers and what has been measured. 6.3.1.3 Currents Currents can be expected to play a major role in determining fish behaviour, but this role is difficult to map out in the eastern Caribbean. Currents in the area have flow rates that vary considerably over space and time, with convoluted meanders and eddy formation being typical (Mahon 1987). Cyclonic and anticyclonic eddies in the island wake may be stationary, or alternately form and shed downstream with a periodicity that depends in part on flow direction and speed. The eddies affect plankton abundance and patchiness through entrapment, and consequently influence fish abundance and distribution. This partial understanding of relationships has not, however, made the spatial or temporal prediction of good catch locations on the basis of currents possible yet. Although flyingfish and other species appear to follow the main circulation patterns at times (perhaps associated with floating debris or river water lenses), counter-current movements have also been reported (FAO 1993). The available scientific data tend to support the high level of uncertainty about currents reported by fishers. But contrary to fishers, marine scientists seem to be more deficient on detailed current observations (particularly about which features are transient and which are more permanent) than on information about the large scale current regime. The scientists are reasonably confident in supplying explanations for much of what is observed, and this may serve to facilitate information exchange between scientists or officials and fishers.  129 6.3.1.4 Migration Understanding fish migration is important to fishers since the small boats have access to the fish only when they move close to the island. This is one of the reasons their fishing season is shorter than that of larger boats. For the larger boats and fishery planners another key issue is the sharing of resources between different national marine jurisdictions. Scientists, however, seem no more informed than fishers since migration has not been mapped out and explained with certainty for any of the pelagic species (Mahon 1987; FAO 1993). Slightly more is known about  the fourwing flyingfish (Oxenford et al. 1993) and dolphin (Oxenford and Hunte 1986) than the other pelagics in the region. These two species appear to have stocks that migrate within the southeastern Caribbean although their distribution beyond the commercially fished areas is unknown. Again perceptions and measurement of uncertainty appear to be similar, especially in terms of appreciation for the deficiencies in basic information. 6.3.1.5 Conservation and management The evidence immediately above suggests that only management on a regional or larger scale is likely to be effective. International management, under ICCAT for example, is accepted as inevitable for the very highly migratory pelagics. However local fishers may also provide valuable input. For example, on the matter of releasing screelers as a conservation measure, the flyingfish scientists agree that this could be advantageous and perhaps the only feasible form of gear regulation. Part of their concern stems from research cruises revealing an apparent scarcity of spawning substrate, leading to new questions about where and on what flyingfish spawn (Oxenford et al. 1993). Although no documentation was found specifically on the integration of ordinary and scientific knowledge as a means of reducing uncertainty, fishery managers and scientists admitted their reliance on the fishing industry for information and supply of affordable research  1,,  1L  platforms. Collaborative work with fishers on various inshore fisheries demonstrated that scientists were willing to take into account ordinary knowledge, and that fishers were willing to play an active part in scientific research and management (e.g., Mahon and Drayton 1992). There is no evidence, however, that this type of collaboration has been utilized in the Barbados pelagic fishery, or that scientific information is being disseminated to resource users by government or academic agencies to educate them about management issues. 6.3.2 Fish prices The interannual and monthly variation in ex-vessel and retail prices from 1981 to 1994 at Oistins fish market is illustrated for the major species in Figure 6.7. A shorter period was available for bilifish. Years cannot be compared directly without adjustment since the time series do not take inflation into account. However, constant dollar comparison is pertinent mainly to long-term perceptions and planning, not routine operations. Some boat owners, for example, have observed that the prices of fishing inputs, particularly gear and equipment, are increasing at a much faster rate than fish prices, thus putting them at a net disadvantage in real terms. Over the 1981 to 1994 period fish prices would have had to rise by around 75% to keep pace with inflation. Clearly this has not happened. Yet, from informal interviews, boat owners’ greatest concerns, and those of the other fisherfolk less concerned with investment, are with the short term, current price, uncertainties over which they hope to have more control. It is on these that I concentrate.  131  Figure 6.7(a) to (e) Variation in fish prices for the major pelagic species at Oistins  (a) Flyingfish ex-vessel  Price (S/kg)  retail  -----•  8.00 6.00 4.00 2.00 0.00 81  82  83  84  85  86  87  88  89  90  91  92  93  94  92  93  94  92  93  94  Year  Dolphin Price (S/kg)  ex-vessel  retail  14.00 12.00 10.00 8.00 6.00 4.00 2.00 0.00 81  82  83  84  85  86  87  88  89  90  91  Year  (c) Kingfish Price (s/kg)  ex-vessel  .-----.  retail  14.00 12.00 10.00 8.00 6.00 4.00 2.00 0.00 81  82  83  84  85  86  87  88 Year  89  90  91  1’.,  Ii  (d) Tuna ex-vessel  Price (S/kg)  retail  12.00 10.00 8.00 6.00 4.00 2.00 0.00 81  82  83  84  85  86  87  88  89  90  91  92  93  94  Year  (e) Bilifish Price (S/kg)  retail  ex-vessel  14.00  12.00 -  10.00  ‘,,  IN  8.00 6.00 4.00  fl’  2.00 0.00 90  91  I  I  92  93  94  Year  Flyingfish is the most variable of the species, and in recent years current prices for this species in particular have been lower than in the past due to low demand rather than high supply. The prices of other pelagic species are more stable, and (on a current price basis) appear to be increasing slightly on average. The most available evidence of lower consumer buying power was found in news media reports. They said that the average Barbadian had less disposable income due either to unemployment in the household or a decision to save more to insure against risks such as future unemployment or currency devaluation. My observation was that consumers were indeed scarce at fish markets compared to  1  ‘•‘  13.)  previous fishing seasons, and purchased fish in smaller quantities. Being traditionally the least expensive fish, whole flyingfish is the choice purchase of lower income consumers in particular. Examination of price records shows, however, that despite lower demand, the differences between ex-vessel and retail prices (the mark-ups) remained similar to those in more prosperous years. I observed vendors unwilling to lower their mark-up in order to sell fish and recoup revenue from increased volume. Instead, unsold fish was stored for redisplay and the volume of ex-vessel purchases was reduced to compensate for inventory, which further reduced ex-vessel prices. Without information on the economics of fish vending it was not possible to adequately explain this behaviour or determine the extent to which it exacerbated the demand and price uncertainty. By comparing the price figures to the landings figures, the expected inverse relationship between price and catch is also apparent. However, since government records do not record catch and price over time during the day, they obscure one of the most critical and dynamic relationships in the fishery. This is the short-term fluctuation of ex-vessel prices in response to anticipated landings during the day, a feature which is especially important in determining the fortunes and behaviour of dayboats. This important contribution to uncertainty has been noted by early and recent authors (Brown 1942; Berkes and Shaw 1986). Daily price variation is most pertinent to flyingfish, the price of which was often observed dockside to fluctuate within a range of $107100 fish during the typical 5 hour window of landings. Most often a decrease over time (not always smooth) was observed, but if the actual landings proved to be less than speculated upon earlier, the ex-vessel price increased for late arrivals. At the minor landing sites the daily fluctuations were smaller and the prices were usually, but not always, on average higher than at the major sites. However it sometimes took longer to dispose of the catch at the smaller sites if few buyers were around. Flyingfish comprise  1—,  IL,  about 65% of dayboat and iceboat landings (Hunte and Oxenford 1986). This observed daily variability due mainly to supply, and the variable overall trend in prices due to consumer demand, create the uncertainty reported by fishers and vendors especially, and recognized by government. The only detailed investigation of the relationships between the landings of pelagics, prices and revenue found evidence of market saturation in which increased landings resulted in reduced total revenue at Oistins (Mahon and Willoughby 1990). Their analyses were largely exploratory and further attention at the level of the individual or enterprise is necessary. They did, however, illustrate that market dynamics are quite complex and poorly understood. This suggests that the fisherfolk engaged in market transactions have reason to be extremely uncertain about price. 6.3.3 Other uncertainties From the planners’ perspective, ice unavailability has a well documented history mainly in Ministry files and the print media. The key points are that the Ministry elected to engage in the provision of ice at its Oistins and Bridgetown markets. But neither has it been able to provide the full quantity demanded (mainly due to fleet expansion), nor has it actively encouraged private investors to supply the ice shortfall. Although constant unsatisfied demand has recently attracted private ice companies, because government ice has a lower (some argue subsidized) price, the harvest sector behaves as reported earlier. A Barbados Development Bank spokesperson surmises that, facing increasing capital costs and fluctuating fish prices, “fishermen have therefore elected to minimize spending and  ...  preferring to buy cheaper ice from  Government [have] created what could be termed a false shortage of ice” since demand exceeds government’s supply (Barbados Advocate 26 Sept. 1994). From interviews, government officials, unlike boat owners, appeared to perceive harvest uncertainty to be uniform because they do not take information sharing into account. Therefore they consider the harvest sector  1,,  ‘3  irrational to accumulate fixed costs and forego income opportunities solely to reduce operating costs. Regarding weather and sea state, examination of daily records revealed that the Meteorological Service in Barbados usually forecasts swells of 1-2 metres which are either moderately increased or decreased by the winds which typically blow from the east. Khokiattiwong (1988), in contrast, found high daily variability and no clear seasonal pattern in wind speed and direction or in swell height. This uncertainty is less important than that surrounding the features of greatest concern to fishers which were the intermittent episodes of northerly swells generated by north Atlantic storms. These high swells affect the Caribbean some time between January to March in most years but are unpredictable in timing and intensity within this window (P. Drakes pers. comm.). These were the months that many east coast boats relocated to fish from Bridgetown. There were several days in the first quarter of 1994 when the dayboats especially were observed to stay in port due to high winds and rough seas. Also, reporting on an extreme episode, a newspaper article stated in January 1990 that fishers had not ventured to sea for two weeks because of high winds. It quotes them as saying such weather is expected due to experience, but if and when it comes is unpredictable (Daily Nation 30 Jan. 1990). But even if the wind does not entirely prevent fishing, by influencing surface currents it can affect catch through decreasing gear efficiency or affecting the formation of the island wake eddies described earlier (Khokiattiwong 1988). Concerning safety at sea, due to increasing mutual assistance rather than reliance on the Coast Guard, recent official statistics no longer reflect the true incidence of fishing boats breaking down. However the mid-1980’s records of the Fisheries Division revealed that boats frequently became disabled at sea due to mechanical or electrical problems so frequently that  136 safety was a constant concern of many fishers and government officials at the time. For example, in 1985 and 1986 about 16% of the fishing fleet required the assistance of Coast Guard, and an unknown additional number of vessels that experienced problems would have been helped by colleagues. While the sea around Barbados is not particularly dangerous on a global basis, and loss of life or limb is rare, the boats are small and ill-equipped. Should fishers drift through the island arc, they are likely to suffer the hardship of not making landfall for several weeks. Apart from the human element, there is also the financial risk to owners. Only a small proportion of the fleet is insured, and recovery of a boat from a distant salvage can be prohibitively expensive if the boat is offered for return at all. Legislation on fishing vessel safety was being finalized during my fieldwork, suggesting that it is still an issue. State intervention is addressed in detail in following chapters. 6.3.4 A note on uncertainty reduction through information exchange  As stated earlier about perceptions of uncertainty about currents, fishers provided detailed observations but seldom seemed to organize or explain them systematically. More information is available on currents than most other oceanographic features of mutual interest to fishery managers and fishers. So a supplementary investigation of fishers’ knowledge was undertaken to assess the potential of information exchange as a means of reducing the uncertainty of both parties. Among the recent literature was a paper, Bowman et al. (1993), that described mesoscale ocean circulation around Barbados in detail. Both the article’s authors and local scientists found the variability and complexity of the current patterns unusual since the findings were not expected on the basis of previous research or prevailing theory. In the first part of the exercise, a group of three experienced fishers was asked, collectively and without prior notice, to describe their observations about the currents around Barbados. There was more agreement than disagreement among them on what they observed. This  137 information, entered onto a rough sketch of the area, was found to conform closely to the findings in the scientific paper in terms of current directions, speeds and seasonal variability. The fishers seemed interested to hear of the scientific findings and to compare them with their own observations. In the second part of the exercise the same fishers were asked to describe observations involving currents which they could not explain. Their observations were entered onto another sketch. One conformed closely to the formation and shedding of eddies in the island wake. When this phenomenon was explained to them they seemed willing to accept the scientific explanation. Afterwards, they expressed interest in gathering other experienced captains and exploring other areas in which fisheries-related scientific information useful to them could be exchanged. This did not occur, but I provided the most enthusiastic fisher with extracts from scientific articles on topics of interest to him and we later discussed the contents. This proved informative to both of us, but communication was difficult due to unfamiliarity with each other’s terminology (a problem also encountered by the Coast Guard during search and rescue). Also, despite limited formal education, this fisher was able to quickly grasp the significance of new information (e.g., types and causes of ocean fronts) fitted to familiar experiences (e.g., conditions favouring successful tuna sets). The experience described above is not necessarily representative of what may occur with most fishers. It implies, however, that limited formal education is not inevitably a barrier to involving the fishing industry in scientific research and management of the pelagic fishery. 6.4 CONCLUDING DISCUSSION Following a brief profile of the respondents, this chapter reported information pertinent to the first research question: what uncertainties are perceived to characterize the Barbados pelagic  1’)  1.,  fishery. Although gender may be an issue in the interface between harvest and postharvest sectors, the most consequential background variables for fishery planning appear to be experience and education. The lack of experience and high levels of formal education of state officials compared to fisherfolk, fishers and vendors in particular, may influence planning perspectives and interaction. Regarding the research question, it is generally perceived by fishery participants and state officials that uncertainty characterizes the fishery. The uncertainties perceived by the fisherfolk in Barbados are similar to the types reported in the general literature (e.g., Gates 1984). Their high levels of perceived uncertainty are in line with the quality and quantity of scientific information available on Barbados and the Caribbean. Fisherfolk, scientists and officials are all constrained by large gaps in basic fishery knowledge. In specific terms, there are features and relationships particular to Barbados and the surrounding marine region. For example, there is a complicated set of relationships between distant freshwater sources, currents, floating debris, fish behaviour and catch. Within the fishing industry there are some, but not many, differences in perceptions of uncertainty according to occupation or technology. The difference between harvest and postharvest sectors on seasonal profitability, and that between dayboat and iceboat owners on input availability and crew turnover are examples. We also see evidence of the three root sources of uncertainty, that is, the marine environment, human activity, and our limits to comprehension. The example above of the series of links between distant rivers and catch illustrates the first, and uncertainty about prices due to the behaviour of fishers and vendors relates to the second. Limits to our comprehension and ability to predict are evident in the lack of a clearly understood relationship between landings and price beyond its obvious inverse nature.  1,,  1.,  Indeed, to answer the first research question directly, catch and price are the maj or uncertainties that characterize the pelagic fishery. Others have been identified, and some, like ice availability and customer demand, may only be transient features, whereas concern about the weather and safety are likely to always be present. Despite the absence of fishery management, state intervention is already a cause of uncertainty. Investigation of the measurements of phenomena perceived to be uncertain showed the causes of uncertainty often to be large gaps in knowledge rather than small doubts about existing knowledge. This was most apparent with almost every aspect of marine scientific information as evident in the multidimensional examination of what factors fishers and scientists had respectively perceived and measured relevant to catch. Here, differences exist in the spatial and temporal scales of observations to which the two have access although the conclusions reached by both parties on some factors, for example, seasonality of species availability, are similar. These differences and similarities have the potential for encouraging dialogue on common ground where information exchange could be beneficial for reducing the uncertainty of both parties. The potential for collaboration in fitting scientific explanations to fishers’ detailed ecological and oceanographic observations holds particular promise. But there are also potential benefits for exchanging information on non-scientific subjects such between the state and iceboat owners on ice supply decision-making. The role of information exchange in planning is a part of a later discussion, and will not be dealt with here further. But a final point of significance is that the uncertainties examined appear to be of such magnitude and prevalence that it would not be unreasonable to expect that social strategies could be employed for coping with them. 6.4.1 Summary • The Barbados pelagic fishery is generally perceived to be characterized by uncertainty. • The most unpredictable aspects of the industry are fish catches, fish prices, customer  140 demand, ice availability, weather and sea state, safety at sea, and state intervention. • Both perceived and measured uncertainties are mainly as large scale gaps in information which may be reduced through information exchange. • Given the magnitude and prevalence of the uncertainties examined, it would not be unreasonable for social strategies to be employed for coping with them.  141 7. SOCIAL STRATEGIES FOR COPING WITH UNCERTAINTY: ATOMISM AND NETWORKS  This chapter starts the investigation of the second research question. It addresses why particular social strategies are used to cope with uncertainties in the pelagic fishery. Evidence relating to the strategies of social atomism and networks is presented here before examining formal organizations in the next chapter. I first look at results from the QFISH and QGOVT surveys which were used, in the context of social network methodology, to design the more detailed SOCNET study reported on next. The primary source of information for this chapter is derived from structured interviews. Additional information from semi-structured, unstructured and informal interviews, as well as from observation and documentary analysis assists in my interpretation of the links between uncertainty and coping. Tables of the results discussed in detail appear in the chapter, with supplementary tabular material provided in Appendix D. 7.1 SURVEY RESULTS I began the fieldwork by using two short surveys, QFISH and QGOVT. These surveys provided basic background data which helped me to focus further research, especially the SOCNET study on social networks in the industry. This first section reports briefly on the survey results. 7.1.1 Income from the fishing industry  Nearly all fisherfolk interviewed (88% to 100%) reported that over 75% of their annual income was from the fishing industry (Table 7.1). In Barbados the fishing industry is the sole economic pursuit of most fisherfolk. The exception to this pattern was owners. Only 35% of owners report deriving a substantial portion of their income solely or largely from the fishery. In  142  fact, 38% of owners report receiving less than 25% of their annual income from the industry. Table 7.1 Proportion of income from fishing industry Proportion of income m fishing industry 9 fr Otol/4 >1/4 to 1/2  .  Fisher (°) 2 3  >1/2to314 >3/4 to all  3 90  Don’t know  2  Owner (°) 38 17  7 35 3  Vendor () 3 6 88 3  Processor ().....  100  Some owners who reported a small proportion of income from the industry said that they were involved in the industry only as a “hobby”, or that they let the boat’s commercial fishing support their recreational fishing, or that they no longer paid much attention to the boat since it had failed to yield much income. The cases in which the owner described the boat as a profitable investment were not common, but this finding seemed to be contradicted by continued fleet expansion. The financial analyses of Burtonboy and Horemans (1988) may provide an explantation for this contradiction and preference for investing in iceboats. Using estimated cost and revenue figures they calculated the return on equity capital (REC) for a typical dayboat owner-fisher or non-fishing owner to be -13% if the boat was financed through a loan, and -6% if the owner supplied all capital. For iceboats the returns were -8.5% and 0% respectively. Although not positive, they showed iceboats to be more profitable than dayboats (recall profits and shares in section 4.2.2). They speculated that positive short-term cash flows could be realized if obligations like loan repayment and insurance were avoided, and depreciation was not taken into account. In the long-term, however, the bank, a disaster or normal wear and tear would erode the unsustainable profit. The above responses of owners suggested that many were facing the long-term reality. These findings also implied that if profit was the aim of boat ownership, then rational means could include highly competitive strategies  1  in order to make money and get out quicidy. Hence the nature of owner involvement in the industry was one of the areas identified for further study. 7.1.2 Occupational alternatives and credit  Examination of occupational alternatives, both in and out of season, and informal credit sources completed the brief economic profile of fisherfolk. Espeut (1992) notes the prevalence of occupational multiplicity among people in the Caribbean with low incomes, including fishers, as a response to uncertainty. He distinguishes between cyclic and concurrent alternative occupations (outside of and within the fishing season respectively). As the income data suggest, 90% of fishers and vendors worked only in the fishing industry during the season (Table 7.2). All fish processors and non-fishing owners had the same jobs in and out of season. In and out of season, only 20% of the non-fishing boat owners depended entirely on their boats for income. The greatest proportion of the remainder were involved in skilled technical and managerial/professional occupations ranging from medical and mechanical services to tourism and engineering. During the off-season 73% of fishers and 91% of vendors still worked in the fishing industry. Fishers mainly switched to other, minor fisheries. Vendors sold demersal, longline or imported fish. Selling fish at Bridgetown tended to be more seasonal than at Oistins since large demersal catches were not usually landed at the former site. As an offseason alternative, 13% of the fishers did skilled manual work such as masonry, carpentry, house painting and automotive mechanics.  144  Table 7.2 Fisherfolk’s alternative work in and out of the fishing season Alternative work in and out of the season No alternative work Fisheries-related Managerial/professional Technician/artisan Commerce/transport Agriculture-related General labour Preacher Administrative Housewife Recreation Sampie(N)  Fisher (%) in / out 88 70 3 3 1 1 3 13 2 3 1 4 1 5 1 1  Owner (%) in and out 20 10 23 20 7 7 3 3 3 2 2  Vendor (%) in / out 91 81 3 10  3  3 3  3  3  Processor (%) in and out 100  L...  -  The state usually puts fishery work and agricultural labour in the same occupational  category (Government of Barbados 1993) and, like agricultural labour, fishing was viewed as an occupation of last resort (Berkes and Shaw 1986). The latter authors say that fishers do not perceive themselves as traditional since recruitment to fishing is primarily due to lack of alternative employment opportunities. There is, however, a strong impression that both harvest and postharvest fishery work is gaining status as a viable alternative to a wider range of occupations for reasons that are not clear, but appear to be both social and economic. Earning supplementary income from small-farming, a commonly cited alternative for fisherfolk around the world (McGoodwin 1990), was not prominent among the respondents. People reported having kitchen gardens mainly for household consumption and limited distribution within a small social circle. Resorting to alternative occupations outside the fishing industry was not a common practice of the fisherfolk sampled, and since the offseason fisheries are minor, reliance on the pelagic fishery is substantial. Credit was a potential economic link between the main and offseason that required  145  investigation. A direct question about fishers credit ties with vendors, suspected to be most common and sensitive, was avoided in the early section of the questionnaire. However, when fishers were asked if they had experienced credit ties to owners just over 10% replied positively. About the same proportion of owners reported credit ties to vendors or processors in exchange for catch. However, nearly 20 % of vendors and processors said they had instituted credit ties with owners or fishers for fish. Being a feature of some importance in the fisheries literature, credit ties were returned to in more detail at several subsequent points in the research. 7.1.3 Kinship  The roles played in the fishery by the kin of fisherfolk are known to vary widely among and within fisheries (Acheson 1981). In British Columbia, for example, around 90% of Indian fishers reported having parents in the industry compared to around 50% among non-Indians (Guppy 1987). In Barbados, while 63% of owners had no kin presently in the fishery, more than 50% of the other fisherfolk did (Table 7.3). Most of these relations were close kin (defined as spouse, father, mother, sister, brother, son or daughter). While the mere existence of kinship ties in the fishery says nothing about the role of kin, there may exist family traditions of occupational inheritance. Other researchers have reported that most Barbadian fishers do not come from families in which fishing is a tradition and they seldom fish with relatives (Berkes and Shaw 1986). The role of kinship in networks was also investigated later. Table 7.3 Involvement of fisherfolk kin in the industry Kin involved n• çlust’ Close kin Distant kin No kin •  .  Sample (N)  Fisher  Owner  (%)  ()  47 9 43  35 2 63  Vendor (%) 59 3 38  •.•J  Processor  () 60 40  .•  146 7.1,4 Degree of fisherfolk interaction  Fisherfolk around the world, especially fishers, have a reputation for uncompromising individualism coupled with the necessity to cooperate in their work (Pollnac 1988a). Because of the conceptual complexity involved, this quick survey was not designed to distinguish between social atomism and individualistic or cooperative networks. It focused just on asking whether fisherfolk coped with uncertainty interactively or non-interactively. The majority of fisherfolk (between 57% and 80%), with vendors the exception (44%), said they coped with uncertainty by interacting with others in the industry rather than by depending solely on themselves (Table 7.4). State officials were asked a similar question in QGOVT on their observations of how fisherfolk coped with uncertainty. All the Fisheries Division and 67% of the Ministry staff thought that it was through interaction with other fisherfolk. The reported lack of instrumentally oriented interaction among vendors may be explained by heightened competition for customers in the fish markets due to the low customer demand noted previously. Table 7.4 Means of coping with fishing industry uncertainty Strategy for coping  Fisher (0%)  Interactive  Non-interactive Sample (N)  Owner Vendor Processor Fish. Div. Mm. Agric. (/) (/‘ (%)  72 28  57 43  44 56  80 20  100  67 33  126  40  32  5  7  4  7.1.5 Social network ties and contents  A ftiller impression of the social relations in the fishery, comes from examining some of the types of instrumental aid most commonly reported in the literature were investigated. These were information on fish catches and fish prices (Tables 7.5 and 7.6), fish marketing (Table 7.7), and cash loans (Tables 7.8 and 7.9). Respondents were asked whether they provided or received the type of aid, and to identify the occupations of their most important network members and  147 their relationship to them. Also, to make sure other support important to the fisherfolk was not missed, they were asked, open-ended, to name the type and sources of aid most important to them (Table 7.10 and 7.11). Network relations are discussed in detail in the following section (SOCNET study), so only a few points which guided the subsequent study, are highlighted here. First, in all questions, over 90% of the respondents reported that the most important people usually involved in either the giving or receiving of aid were not related to them. This suggested, subject to further investigation, that kinship did not play a very important role in exchange of instrumental aid. This is an important finding, one that is contrary to the situation in many other small-scale fisheries although not completely unheard of (Acheson 1981). Second, the importance of network member occupations differed with the type of aid received or provided (e.g., owners asked more fishers about catch than price), and exchanges were often asymmetric. That is, more and/or different aid was received than provided (e.g., all processors received price information, but not all provided it). Other impressions were that fishers tended to interact more amongst themselves than did other fisherfolk, and that while credit ties were important, they were likely to be underreported. The results of the open ended question suggested that the four main aid types being investigated, and particularly information on catches, were utilized. However, the low ranking of marketing arrangements and information on prices, and the higher ranking of financial assistance among vendors compared to other occupations, required particular attention. Other useful types of aid were search and rescue, and fishing or seamanship advice. From comments made during interviews, however, the impression was given that the latter was limited to dyadic ties between fishers only. Hence this was less relevant to the type of network and organizational study being conducted here. The importance of boat maintenance and repair to owners, and general  148 information on activity in the fishery to processors, were also noted. The large proportion of vendors who claimed not to know the type of aid most useful to them was taken as a hint that, having apparently not given it much prior thought, they perhaps had less need of assistance.  Table 7.5 Catch information networks Catch info. network members Fishers Owners Vendors Processors Others Sample (N)  Rec. (%)  Fisher Prov. (%)  100  100  122  121  Owner  Vendor  Processor  Rec. (%)  Prov. (%)  Rec. (%)  Prov. (%)  97 3  81 16 3  76  61  19  33  5  6  25  21  18  4  32  32  Rec. (%)  Prov. (%)  75  67 33  3  Table 7.6 Price information networks Price info. network  Fisher Rec. Prov.  Owner Rec. Prov.  Vendor Rec. Prov.  Processor Rec. Prov.  members  (%)  (%)  (%)  (%)  (%)  (%)  (%)  Fishers Owners Vendors Processors Others  80 2 8 2 8  96 2 1 1  56 9 29  74 17 9  31 8 61  28 4 68  Sample (N)  96  95  35  26  25  6 34  (%)  40 40 20  67 33  5  3  Table 7.7 Marketing networks Market prop.  network members Fishers Owners Vendors Processors Others Sample (N)  Fisher  Rec. (%)  47 53  19  Prov. (%)  61 39  26  Owner  Rec. (%)  40 60  10  Vendor  Prov. (%)  Processor  Rec. (%  Prov. (%)  36 64  56 44  62 38  8  11  9  Rec. (%)  Prov. (%)  60  75  20 20  25  5  4  149  Table 7.8 Small loan networks Small Joan network  members  Fisher Rec. Prov.  Owner Proy. Rec.  (%)  (%)  (%)  Fishers Owners Vendors Processors Others  25 63 4 4 4  75 7 14 4  50 50  Sample (N)  24  28  2  Vendor Prov. Rec.  Processor Rec. Prov.  (%)  (%)  (%)  (%)  94 6  25 75  22 45 33  18  4  9  (%) 50 50  0  4  Table 7.9 Large loan networks Large loan network  members  Fisher Rec. Prov.  Owner Prov. Rec.  (%)  (%)  (%)  Fishers Owners Vendors Processors Others  30 40  65 17 18  Sample (N)  10  17  Vendor Rec. Prov.  Processor Rec. Prov.  (%)  (%)  (%)  (%)  100  71 29  25 75  14 57 29  1  7  4  7  (%) 50 50  10 20 0  4  Table 7.10 Most useful types of aid received  Most useful type of aid Fishing, seamanship advice Search and rescue Information on fish catches Bait and input supply General fishery infonnation Boat maintenance, repair Boat launch, haul-up Financial assistance Fish handling advice Fish marketing proposals Information on fish prices Other types of aid Don’t know Sample(N)  Fisher  Owner  Vendor  (%)  (%)  (%)  18 17 15 7 7 7 5 1 1  Processor 25  17 19 11 30 6  5 9  50  27 9 5 25  10 12  3 14  9 36  113  36  22  4  150 Table 7.11 Sources of most useful types of aid received Network •. membe Fishers Owners Vendors Processors Others  Fisher  Owner (°4).. (°) 81 80 1 3 1 10 10 7  7  Vendor (  Processor 50  33 33  25  27  25  7  Sample 7.1.6 Power To measure perceptions of power, respondents, including state officials, were asked to name who had the most influence on how the fishing industry operates (Table 7.12). The processors were seen as the most influential by all categories except the fishers who identified the vendors. Presumably this was partly because fishers deal more with vendors than processors. Processors also ranked themselves as having the most influence. This strongly suggests that the postharvest sector wields the most power in the industry. Boat owners were clearly perceived to have the least influence. Government was most often ranked third or a tied second. These results were important for guiding further investigation given the centrality of power in social relations and organizations. It appears that there is a perceived hierarchy in the industry with the state subordinate to the elite in the postharvest sector. For almost everyone the postharvest sector wields the most power in the industry. Evidence of power dynamics and potential power in networks were key issues for further study. Table 7.12 Who has the most iiffluence on how the fishing industry operates Category with en Fishers Owners Vendors  Fisher  •  Government Don’t know  8 5 46 26 10 6  Sample (N)  125  Processors  Owner Vendor Processor Fish. Div. Mm. Agric. (%) (%) 8 26 25 17 3 17 28 10 14 17 45 39 50 71 33 13 10 25 14 17 5 16 40  31  4  7  6  151 7.1.7 Desired industry improvements With no goals or choices of direction, there is little point in devising strategies to cope with uncertainty. However, if there are goals, such as improvements people would like to see in the industry, then one could expect them to be related to major uncertainties. And, one could also expect people to seek improvements either through their own efforts or those of others, depending upon who they thought was capable and should be responsible for making such improvements. Pertinent to this and subsequent chapters, respondents were asked what was the most important thing they would like to see done to improve the fishing industry for themselves (Table 7.13), and who should be responsible for making the improvements (Table 7.14). The results reveal a reasonable fit between what people found to be most uncertain and what they wanted done to improve things. We see that fish price stability and related state interventions into the market (e.g., state marketing and processing, and protection from imports) are more prominent in the harvest than postharvest sectors. Credit is mentioned, but the only reference to catch is a desired increase in amount, not stability, from a processor. The conspicuous absence of catch stability or predictability suggests that these improvements are perhaps inconceivable or not feasible to respondents. The relatively high ranking, by all respondents except processors, of the need for cooperation among fisherfolk signals the need for a very close look at conflict and organizational factors.  In the majority of cases (65% to 100%) respondents said that both the fisherfolk and the government should be responsible for improvements. If fisherfolk want both themselves and government to assist in improving the industry, then we should see evidence of strategies employed to do this both by self-help and by persuading government to take favourable action. Yet not all the government officers’ desired objectives coincide with the fishing industry. Conspicuously absent are price and marketing.  152  Furthermore, 50% to 100% of respondents in all categories perceived difficulty in the state and fishing industry working together to achieve objectives. Processors and the state were unanimous on this point (Table 7.15). If fishers and vendors had not perceived the researcher as being connected to the state, perhaps the “yes” response from them would have been higher. The reasons suggested for the perceived difficulty (Table 7.16) also assisted in informing the examination of social networks and organizations, and are returned to in the later chapters. Perceptions of industry conflict, state indifference and poor communication are widely shared, but lack of education among fisherfolk and state bureaucracy are also perceived to be barriers by the Ministry and processors respectively. These differences and similarities further emphasize the need to examine relations within the fishing industry, and the relationship between the fishing industry and state to comprehend events and behaviour in the fishery. The social relations and organizational aspects are discussed below and in the next chapter. Table 7.13 Most important desired improvements to the fishing industry Most important fishing industry improvement Fish price stability Physical infrastructure Increased ice supply Fisherfolk cooperation Cheaper fishery inputs State marketing, processing  Fisher  Owner  30 19 11 10 8 7  13 15 20 10 13 15  Easier access to state credit  5  5  Stop fish importation Safety services, training  2 2 1 1 1  Bigger fish catches  Reduce bureaucracy Regional fishing access  Vendor  Processor  3 35 13 19 3 3  20  *  Mi Agric.  20  14 14 14  50 33  20  6 3  20  Reduce foreign fishing Better fish quality Better state/industry dialog Better scientific data Proactive fish management Donotknow  Fish. Div.  (.°‘)  20 14 14 14 14 3  .5 ) This column totals 98% due to rounding.  6  18  17  1  r  I .)L  Table 7.14 Who should improve the fishing industry Who should improve thefishingindustry Government alone Industry alone Both Don’t know  Fisher  Owner  Vendor  Processor  Fish. Div.  (%)  (%)  (%)  (%)  (°4)  26 7 65 1  33 3 65  19 6 72 3  100  100  Mi  Agric.  100  ...................  Table 7.15 Perception of difficulty in state/industry cooperation Is there difficulty in state/industry cooperation Yes No Don’t know  Fisher  Owner  Vendor  Processor  Fish. Div.  (%)  (%)  (%)  (%)  (%)  51 41 8  73 27  50 44 6  100  100  Mi  Agric.  (%) 100  Table 7.16 Reasons for difficulty in state/industry cooperation Reasons for difficulty in ii4qslry  ptic1  Fisherfolk conflict State indifference Poor conmunication State corrupt, unreliable State lacks finances Fisherfollc not educated State bureaucracy Fisherfolk unreliable State lacks fishery data Inadequate infrastructure Don’t know  Fisher  Owner  (°4)  (°4)  28 27 20 15 6 2  32 11 25 11 4  Vendor  Processor  13 7 40 13 13  Fish. Div.  Mi  (.°4’)  ......  20 20  58 14  60  14  Agric.  (2) 17 17  66 7 4 7 14  2  6  7  ,  7.2 SOCIAL NETWORK GENERAL FEATURES  The results in this and the following sections derive from the SOCNET study of 37 respondents,  supplemented by unstructured  and informal interviews,  documents  and  observation. The overview of general features provides a brief look at network structures before examining relations in detail from the perspectives of people in each respondent category. The network terms introduced in the literature review are used to describe ties between the focal  154 individual (ego) and network members (alters). To reiterate, in this study, the theoretically relevant content of a tie is usually some sort of instrumental aid (e.g., information). But a tie can contain more than one type of aid (e.g., credit and price information) in which case the tie is ‘multistranded’, each type of aid comprising a single ‘strand’. The other terms used have common usage, except ‘links’ which refer to relations between alters rather than between the ego and alters (see Figure 3.3b). In this study, at least half of the respondents in all categories (50% to 77%) said that both they and others coped with the “ups and downs” of the industry by interaction rather than on their own (Tables 7.17 and 7.18). Allowing for limitations in the survey method, these results agree reasonably well with the QFISH self-reports (Table 7.4). Fishers saw themselves as employing interaction more than other people in the industry, whereas owners and especially vendors placed others as being more interactive than themselves. Processors’ responses were split between doing things mostly on their own and interactively in both self-report and reporting on others. Most respondents said that, from an economic standpoint, interactive strategies were a necessity since, without “connections”, you “couldn’t last long” in the industry. Some added that sharing information was useful to know how “to work to suit”. The reasons for non-interaction were vague, most respondents just citing avoidance of conflict in the industry. Probing revealed that non-interaction was not synonymous with social atomism since respondents reported that ties existed but were used only sparingly, not routinely, so as not to accumulate obligations. Credit ties were especially perceived as almost inevitably conflictual obligations. The perception of fishers that they are highly interactive is consistent with data on their behaviour at sea.  155  Table 7.17 Respondents’ strategy for coping with uncertainty Respondents’ own strategy for copin Interactive Non-interactive  Fisher  Owner  Vendor  (0/)  77 23  Processor  () 50 50  57 43  50 50  Sample  Table 7.18 Perceptions of others’ strategies for coping with uncertainty Respondents perceptions of others’ strategies for coping Interactive Non-interactive Do not know Sample (N)  Fisher  Owner  (%)  (%)  50 44 6  60 30 10  18  10  Vendor (.°<) 71 29  7  Processor  (°). 50 50  2  7.2.1 Aid utilized  In terms of how the strands in respondents’ networks contributed to aid, the largest contribution in types  all cases was the diverse “other aid”. This consisted of aid in addition to the 5  specifically identified (Table 7.19). This is discussed later, but in few instances did any  single component exceed any of the specific aid types. Regarding the aid types of primary interest, fishers and owners followed the same pattern with most strands providing information on catch, then prices, then marketing proposals. The strands of vendors and processors mainly concerned information on prices, but vendors listed marketing proposals next, whereas processors followed first with information on catch. Catch information was little used by vendors, but presumably concerned processors more because of the greater volumes of fish dealt with. Strands supplying credit were reportedly few in all cases. More strands were reported for the short term or chronic concerns of each occupation than longer term or occasional requirements.  156 Table 7.19 Fisherfolk’s average network strands of aid Strands containing types of aid as Fisher Owner Vendor Processor (%) % of total network aid content (%) (%) (%) 23 7 Information on catch 19 18 15 Information on price 16 20 20 Marketing arrangements 13 11 19 17 11 8 11 11 Loanof$10-$500 Loanof $500-$5000 Otheraid  7 30  Sample (N )  18  4 40  6 40  5 29  7.2.1.1 Average network composition In terms of who were the most common network members for supplying aid in general, fishers and vendors in the sample each have members of their occupation comprising the greatest percentage of their networks (Table 7.20). Boat owners list fishers as their main component, and processors list vendors as theirs. The other fishery occupations in owners’ networks are mostly boat agents listed by Bridgetown respondents. Processors appear not to supply each other with any type of aid, but of the fisherfolk they have the highest proportion of contacts with government and non-fishery occupations. The limited representation of government employees  in the networks of the other fisherfolk, especially fishers, supported the earlier findings that perceptions of bureaucracy are related to levels of interaction with the state. Table 7. 20 Fisherfolk’s average network composition by member category Member category as  Fisher Owner Vendor Processor  % of total network Fishers Boat owners Fish vendors  (%) 40 18 21  Fish processors Government employees Other fishery occupations Non-fishery occupations  6 2 6 7  Sample (N)  18  (%) 35  (%)  (%)  10  25 17 38 1 6 8 5  12 6 20  10  7  2  13 21 7 3 11  15 18 29  A high degree of interaction within an occupational category may be symptomatic of close knit networks based on survival-oriented ties (Rodman 1971), and access to a wider range of  157  resources from more distant ties may be constrained (Granovetter 1973). 7.2.1.2 Ties, strands, gender and relationships Looking at Tables 7.21 and 7.22, it appears that the processors have more ties and more complex relationships with their network members than the other fisherfolk. Multistranded ties are reported to be more durable and constraining in the sense of shaping action and limiting the formation of new ties (Granovetter 1973). Presumably, whether multistranded ties have more effect on the ego or alters may be determined by relative power in their relations. This is pursued in the detailed study of processors. Gender composition of networks is fairly closely linked to that of the industry in general. This is shown by women being in greatest proportion amongst the vendors especially since, as shown above, they tend to deal the most amongst themselves. Kinship in networks is low, but highest amongst vendors, and only fishers and owners have appreciably more friends than associates in their networks. The higher proportion of kin in vendors’ networks may reflect the finding by Sutton and Makiesky-Barrow (1981) that women are more active than men in maintaining kinship networks for support. Given that this support is often for child-rearing (Powell 1982), it may be better developed among the vendors at Oistins where kin tend to live in close proximity. It would be useful in future gender studies to determine if such domestic networks are translated into commercial ties between vendors since this may assist in explaining the dominance of females at Oistins especially.  158  Table 7.21 Fisherfolk’s average network composition by ties, gender and affiliation Member ties, strands, gender and relationship with focal individual Number of ties or members (n) Number of strands (n) Numberofstrandspertie(nln)  Fisher Owner Vendor Processor 15 21 1.4  15 22 1.5  15 20 1.3  17 29 1.7  Male members (%) Female members (%)  81 19  89 11  65 35  82 18  Members who are kin (%) Members who are friends (%) Members who are neither friends nor kin  6 61 33  6 59 35  10 44 46  50 50  (%)  7  18  ..?  Table 7.22 Multistranded ties (number of strands per tie) Network members Fishers Owners Vendors Processors Other fishery Government  Fishers 1.4 1.4 1.3 1.3 1.3 1.0  2.0  Owners 1.3 1.3 1.7 2.2 1.2 1.0  Vendors 1.4 1.4 1.8 1.0 1.0 1.0  Processors 1.6 1.7 2.3 —  1.5 1.0  .:?  Table 7.23 Proportion of friends in the fishing industry Proportion of friends iithe 4ustiy  From0to25% Over 25 to 50% Over 50 to 75% Over 75 to 100% m  ..  Fisher (°) 11 39 28 22  Owner  Vendor  Processor  .(°‘‘)  ()  (°4)  60 30  50 50  10  29 14 43 14  10  7  ..  Given the previously mentioned theoretical significance of network composition in terms of access to distant sources, respondents were also asked the broader question of what proportion of their friends worked in the industry (recognizing the limitation of “friend” being interpreted differently). At the 75-100% level, fishers have the greatest proportion of friends in the  159 industry, but their modal response was at a lower proportion than vendors (Table 7.23). It was frequently reported by fisherfolk that “the majority [of fishers and vendorsj still hang out amongst themselves” and they “still cut away from the general public” (SOCNET Interview No. 11). Processors and owners reported the lowest proportion of friends in the industry. 7.2.1.3 Linkage Also related to the previous questions, respondents were asked to estimate what proportion of their network members interacted amongst themselves (Table 7.24). This was so as to roughly determine whether their network could be considered closely knit or if there was truly a network structure rather than disconnected dyadic ties. Only in the case of processors was linkage among members particularly low due to the high proportion of non-fishing alters. Vendors reported the highest linkage. A high degree of linkage does not necessarily mean that instrumental aid is being exchanged, but it implies that network members potentially have access to more social capital than if the ties were purely dyadic (unless the number of dyadic ties was very large). Persistent, strong links can give rise to cliques, highly socially cohesive groups, which may form the basis of collective action (Wellman 1988). Table 7.24 Linkage of network members Proportion of linked network members Few  About half Most  Fisher  Owner  Vendor  Processor  (%)  (%)  (%)  (%)  14  50 50  11 11 78  30 70  86  There was also a high degree of overlap between the networks of Oistins and Bridgetown fisherfolk although respondents were selected individually, not by snowball sampling. Although flows of resources along links were not measured, these findings suggest that the predominantly dyadic ties found in previous periods and other Caribbean societies by Mintz (1971) and Gussler  160  (1980) may not be typical of the Barbados fishery. Specific investigation of linkages, not included in this study, would be required to properly address this point. 7.2.1.4 Reciprocity and conflict Perhaps from being towards the end of the interview, or because of sensitivity, the responses to questions on reciprocity and conflict were vague and of fairly poor quality. However, the predominance of general over specific reciprocity is clear in all cases (Table 7.25). Taking into account all other evidence, particularly on credit, the reported incidence of conflictual ties from this data source is unrealistically low. Besides fatigue and sensitivity, another possible reason for few reports of conflict is that aggressive business practices may not be interpreted as conflictual in a competitive market economy. The processors most readily admitted that competition could often be “rough”. Table 7.25 Fisherfolk’s average network by extent of reciprocal and conflictual ties Reciprocal and conflictual ties Fisher Owner Vendor Processor as % of total network (%) (%) (%) (%) 24 Specific reciprocity 17 21 18 72 General reciprocity 73 66 82 Absence of reciprocity 4 10 13  Conflictual Non-coifflictual  5 95  4 96  3 97  100  Other studies (e.g., Wellman et al. 1988) have also found that general reciprocity tends to predominate in personal networks, and that network exchanges tend to be asymmetric. While the latter point is not presented numerically, several respondents identified ties in which they either provided or received a disproportionate amount of assistance. The few cases of conflict represent those in which the respondents found the imbalance too much to their disadvantage. 7.3 FIsHING INDUSTRY NETWORK DETAILS  Discussion in this section continues within the context of the general conceptual model  161 (Figure 3.1) from the perspective, in turn, of each of the fishing industry respondent categories as focal individuals. Besides reporting on specific network relations, additional general factors that shape the perceptions and relations of fisherfolk, and assist in explaining why particular strategies are chosen, are examined. Only points upon which there was substantial agreement or support from other evidence are discussed, but important points upon which there was disagreement or insufficient evidence are also mentioned. There is unavoidable overlap among sub-sections, but this facilitates comparison and contrast. Individualistic and cooperative network strategies are also differentiated. Comparison and contrast between Oistins and Bridgetown is made where the data permit. The tables are based on the means of values for the respondents in each of the four occupational categories in the SOCNET study. 7.3.1 Fishers as focal individuals Table 7.26 Aid received from the network of a fisher Fishers’ network members  Information on fish catch  Information on fish price  Marketing proposal  (%) Fishers Owners Vendors Processors Other fishery Government  () 87 10 1 2  44 12 35 2 8  2 75 19 4  Loan of $10 - $500 (.°<) 16 18 22 6 2  Loan of $500 $5000 -  Other types of fishery aid  () 17 28 10 3  39 19 6 6 8 7  .‘  7.3.1.1 Fish catches The most common strategy of fishers for coping with catch uncertainty was to obtain information about factors related to the location of good catches. There were several ways for fishers to do this, and the anticipated outcomes were to reduce search times, and increase catch or catch rate with consequent increased earnings. Not unexpectedly, the average fisher in the sample received most (87%) of the catch  162  information from peers, followed by the boat owner (10%) (Table 7.26). Regarding the nature of the relationships with network members who provided this type of aid, fishers were about evenly divided between long-term friendships that extended beyond providing this type of information, and non-friendship ties on the basis of good access to information and with a history of truthfulness. At the landing site, some fishers obtained information from the owner of the boat on the identities of successful fishers who had landed earlier and could be asked to “give the course”. Since the owner was invariably at the site to oversee fish sale or check on the boat, functions not always well tolerated by fishers, this type of aid was volunteered or solicited only in cases of good relations between the fishers and owners (mainly of dayboats). Iceboat fishers sometimes received information at sea from boat owners who operated base stations and had access to radio communications or other information not received by the boat. Among fishers ashore, more information was reportedly exchanged around the immediate landing site than elsewhere due to the absence, at most places, of the fishing villages typical in accounts of small-scale or artisanal fisheries. Discussion groups typically comprised 3 to 7 persons who fished from the same site, but not necessarily as a group. Unlike the case of owners, the information exchanged here usually concerned most of the factors involved in the search for fish, either observed directly or obtained from the radio at sea. In the latter instances, debates about the reliability of information occurred depending on personal acquaintance with the source and corroborating evidence. Also ashore, at night fishers telephoned friends who fished from other bays in order to compare information. For example, there were phone calls between Oistins and Conset Bay fishers since the fleets often fished different areas, but with some overlap. This level of communication appeared to be mostly, but not exclusively, between owner fishers.  1  103  At sea, communication was more complex. Among dayboats radio communication was virtually continuous and seldom limited to others from the fisher’s home port. Some fishers used simple radio code, for example, “I going up home”, to unobtrusively signal a fishing partner to switch to a pre-arranged higher frequency for a brief secret exchange. No evidence of the elaborate codes and code groups reported in other fisheries were found (e.g. Stuster 1978). Since few frequencies were used and the codes were well known, this ruse offered only limited or temporary secrecy. Codes were said by some to be shared more among fishers from a landing site than between landing sites, but there was insufficient evidence to confirm this. More often there was deception, not so much about location and whether fish were being caught, but about the catch and catch rate which were usually reduced rather than exaggerated. Some fishers said that they never asked for information, but just listened to others on the radio and made their own judgments, an individualistic strategy, free-riding on the networks of others. Palmer (1990) describes similar behaviour off Maine as a common short-term economic strategy. Barbados fishers explained that the resulting “problem” of owing someone a favour, and the likelihood of deception, often outweighed the benefits if you called someone other than a close friend. Some fishers were said to brag about successful deception, but these were in the minority. Fishers explained the non-malicious reasons for deception as just part of the competition to show who was best without assistance, the protection of expected income if you thought you were doing well on a generally poor day, and the related need to prevent accurate catch information from reaching the networks of vendors. Respondents agreed that iceboats, more often than dayboats, called their colleagues “for the drift” (typically floating tree branches or debris with associated schools of fish). They did this when they were ready to head home, and it virtually guaranteed the other iceboat(s) a high catch rate. Iceboat fishers explained this information sharing in terms of being less congested and  164 competitive than dayboats because of their wider search areas. But they also said that most iceboats keep in touch regularly anyway by radio for safety purposes due to the distance of their fishing grounds from home. This set up a norm of cooperation. Although not stated explicitly, it was implied that supplying rather than withholding fish catch information was insurance for a better search and rescue (SAR) response from colleagues if ever required. An unintended outcome of the efficiency of iceboat cooperation was found in the complaints of dayboat fishers that iceboats, particularly when several fished close to the island, “cut off the fish” and “fished out the wood” up-current, increasing dayboat uncertainty of finding fish around the floating objects. Indirect conflict due to interception fishing has been noted for Barbados previously (Berkes and Shaw 1986), and is common in other places including the east and west coasts of Canada (Pearse 1982). The latter, however, tend to be conflicts between gear types. In Barbados, since the boats use the same gear, it is more a matter of scale and efficiency. Similar problems exist in southeast Asian fisheries undergoing modernization (Thomson 1980), but the Barbados situation is mild in comparison. Among other things, the difference in vessel scale is relatively small and not all species are affected to the same extent. When fish, particularly dolphin, was found, the dayboats tended to be secretive or deceptive, pointedly ignoring their normal network obligations by not cooperating to reduce the uncertainty of their colleagues. Some allegedly removed objects or tried to sink them after fishing so that their competitors in the fiercer “fresh fish” market could not lower prices with subsequent large catches. This individualistic strategy was used since the next day often started at the previous day’s prices. Another individualistic strategy with the same goal was to temporarily move away from an object to mislead other boats who were searching and could be tempted to investigate a boat that had stopped to fish. At Oistins, one dayboat fisher, who for a  165  few days was consistently catching more dolphin than others, was said to be “tricking the fellows” by putting to sea much earlier and secretly so that no one could follow him to his unknown fishing area. He also kept off the radio for these days in order to have no obligation to share information. The other fishers who considered him to be in their network apparently did not object to this breaking of ties for individual achievement since they said that they would do the same in that situation. Because the boats had different ranges, and were thought to sequentially fish in the same bodies of water brought closer to the island by the currents, some fishers said that they especially sought information from network members in the larger boat types so as to get advanced knowledge of what to expect in their area in the next few days. This information transfer was least conflictual from longliners to iceboats since they targeted different species, but was also common from iceboats to dayboats since if the former followed fish into “dayboat territory”, and fished among dayboats, they stood the chance of ridicule. There appeared to be a norm of spatial fleet separation which reduced the potential for direct physical and economic conflict due to congestion. On the fishing grounds you were obliged to “give the other fellow a chance.” This is conceptually similar to the more formal rights to access and territoriality common to inshore demersal fisheries worldwide (Acheson 1981). But since both the pelagic fleets and resources are highly mobile, it is much more informal and flexible. Uncertainty was also reduced by egalitarian relations between captain and crew. While the captain was expected to suggest the fishing location, the final decision was often by consensus except on longliners where the fishing was more specialized and dependent on instrumentation. On dayboats and iceboats, the one or two other fishers onboard usually supplied their own information and suggestions for decision-making to an extent that seemed to vary markedly between boats. Since it was the captain, not the owner, that chose the crew in almost all cases,  166 he set the criteria for their selection. It was said that that iceboat crew were chosen more on the basis of seamanship or fishing skills (that presumably could reduce uncertainty), and that dayboat crew were chosen more on the basis of casual friendship with the captain (with less obvious contribution to uncertainty reduction). 7.3.1.2 Fish prices, marketing and credit The social relations involving fishers in pricing, marketing and credit are especially complex since, as one fisher put it, “everything is connected” (SOCNET 14). Obtaining information on fish prices is a strategy used to choose between ex-vessel sales options in both the open market and in marketing arrangements, since the latter do not include contractual prices and are flexible. However, a fisher’s credit and other ties to those in the postharvest sector influence both price and marketing arrangements, which affect access to further credit. The fishers’ preferred outcome is apparently continuity of income rather than either the short or long term maximization of earnings. It is in this intricate web of relations that individualistic strategies and conflicts of interest are most evident. Although other fishers mainly supplied information on prices (44%), vendors were also important (3 5%), and in marketing arrangements they occupied primacy of place (75%) ahead of processors (19%). Sources outside the fishing industry were the main providers of both small and large loans (35% and 41% respectively). A fair amount of credit was also obtained within the industry except from the processors with whom fishers seldom interacted directly. There was a tendency to borrow larger lump sum amounts from owners (28%) than vendors (10%). In selecting network members for fish price information, fishers focused more on reliability (60%) than friendship (20%). The modal reason for choosing a particular person for a marketing arrangement was split between the person being able to buy large quantities of fish, and their ability to agree on a fair price (around 40% each). About 15% of fishers had their marketing  167  person referred to them, often by the owner. Significantly, no fishers said that their marketing alters were chosen due to credit ties, although this seems frequently to be the case. Fishers chose mainly friends and kin as the sources of small and large loans (about 65%), followed by people who were neither of the above, but with whom they frequently did business, such as owners and vendors. Once ashore, one overhears the prevailing prices within a few minutes of being at a landing site, so there is seldom the necessity to ask. If they do ask the going price, fishers said they preferred to rely on their peers since they cannot trust vendors to always tell them the truth. Yet a dayboat fisher will often come ashore and immediately ask of no one in particular, or the vendor with whom he has a marketing arrangement, “How the fish selling?”, and accept whatever response is heard unless others immediately say otherwise. If the prices have been volatile (only with flyingfish) another fisher may occasionally offer advice on the current trend. In contrast, iceboat fishers only proceeded with fuller information, usually waiting until the owner arrived or until the next day before selling. When several iceboats were selling, they consulted amongst themselves before and during sale. At sea, dayboat fishers, generally being price-takers, did not exchange price information much. Exchange by radio was more common between iceboats in port and those at sea, since the latter have some scope to schedule their return for favourable prices. Many said, however, that often by the time they landed the price had changed to the extent that the information was not much help. This was usually because several boats had received the same price information and chose, individualistically, to land at the same time. Cooperation on information did not extend t