A GUIDE FOR THE SMALL-SCALE FISHERY ADMINISTRATOR: Information From The Harvest Sector

...... " , ., .... ,. :>,' - o . ..

DAVID STEVENSON RICHARD POLLNAC PHILIP LOGAN

ICMRD

International Center for Marine Resource Development 1982 University of Rhode Island, Kingston, Rhode Island

A GUIDE FOR THE SMALL-SCALE FISHERY ADMINISTRATOR: Information From The Harvest Sector

DAVID STEVENSON RICHARD POLLNAC PHILIP LOGAN

ICMRD 81/82.005

ItWDati..u Cate fr Marine Re ouze Developmen 1982 Unlvers of Rhode I d, Kinston, Rhode Isand ACKNOWLEDGEMENIS We would like to express our appreciation to Sandra Prata for endless typing and retyping of the original manuscript, to Laura Borasio for her patience with the extensive changes in the final manuscript, to Nan Garrett for the computer generated figures, and to our colleaguesi W. Kotberg and J. Johnson for their early contribu­ tions to this guide. Finally we would like to express our appreciation to our colleagues who work in fishery development in Panama, Costs Rica, El Salvador, and Guatemala, countries in which the authors have refined their ideas conceming data needs for fisheries development. While this guide was financed by USAID (Grant No. AID/DSAN-G-0173) the find. ings and conclusions are solely those ofthe authors. PREFACE actually collecting the data, and rec­ This guide was prepared as a general collect, for individuals in devel- ognizing what constitutes data, and 4) the resource document level of specificity in identifying data which oping countries who either require or are to another. information will be valid from one fishery responsible for providing authors' experience, fishery for making policy decisions In the which is used office personnel are more likely to be trained which affect small-scale fisheries. Its pur- economists; in identify- as biologists rather than pose is a practical one: to assist and sociologists with expe­ data and in designing and anthropologists ing important extremely rare. A fair data collection programs which rience i fisheries are required executing the amount of professional training is will generate information concerning into useful iformation in sector of any small- to convert data resource and harvesting each field. Thle analyses used in economics scale fishery. and sociology/anthropology, while they are The first objective of this guide is to are appli­ eco- evolving, are fairly standard And describe a minimum set of biological, in countries in differ­ information from cable to studies made nomic and sociocultural stages of development. This is less so in small-scale fishery resource and har- ent the so the case of resource assessment analyses; vesting sector which should be available is devoted to describ­ policy decisions can be thus, some attention that intelligent ing the models used in these assessments. made. The first three chapters are directed the describes On the other hand, it is easier to specify at this objective. The first chapter economic data between the fishery primary biological and the interrelationships for analyses. Sociocultural data the economy and stresses the impor- needed and needs are much more specific to given sites, tance of adequate information for decision and cultures.* of sociocultural communities making. The importance to actual data collection, is emphasized in Chapter II. As we proceed information it is only in the field of economics that a The third chapter outlines information formulated (at decisions in the basic set of questions can be needed for effective policy and applied with some modifi­ and harvesting sector of the this distance) resource cations, by the non-professional staff fishery. member. Specific measurements and obser­ The second major objective of the guide by the fisheries data collection vations can be described is to describe appropriate but, when species identification is will generate the needed biologist, methods which required, that expertise must be present dur­ information. The final three chapters of the as well as ing data collection. guide are directed at this goal althcugh the guide gives the recommendations for Finally, towards developing impression that a seemingly endless multidisciplinary data collecL strategies is required, These chap',ers were amount of data and information in order to reduce costs. this is not the case. It is true that much more intended primarily to assist fishery officers than has histori­ designing and information is required who are responsible for appreciated. The justification for data collection process. Data cally been directing the a holistic approach, however, is presented needs are outlined in Chapter IV, collection and the economies the organization in Chapters I, II, and III methods in Chapter V and derived from a coordinated of data collection methods which can be and integration approach to data collection are outlined in in Chapter VI. Finally, the guide includes a considerable information and ater VI. Although lists of background of information is required, a ratio­ data acquisition forms in four amount selected nale has been provided for its collection.** appendices. use of hand calculators and words should be said about the The increasing A few mini-computers means that the storage, different emphases placed on the biological, of the retrieval and manipulation of large quanti­ sociocultural and economic sections accomplished in to which each ties of data can soon be various ciapters. The extent countries. Improvements discipli:±cs is discussed and even the poorest of these three in data processing facilities and a growing the nature of the discussions themselves of small-scale discipline, i.e., awareness of the importance reflect: 1) the nature of the in many developing countries examine, 2) the probable fisheries what it seeks to make it more important now than ever of a fisheries office, 3) the profes- needed infor­ staffing as before to acquire this urgently sional training required for tasks such it to promote rational which data to mation and to use analyzing data, indentifying development and management programs. economic data and analyses with regard to MEY, in spite of some popular misconceptions concerning the role of *This holds will be said about this at the end of Chapter IV. MSY, and especially OSY (Optimal Sustainable Yield). More attractive research and dissertation material for this information is rational and interrelated will make it ** The fact that other countries; hence, it is much less likely to go unanalysed. university faculty and students from both the involved and ii CONTENTS FOR DECISIONS 1 I. THE FISHERY AND THE ECONOMY: INFORMATION 1 System ...... 1.1 The Fishery 1 1.2 The Small Scale Fishery ...... and harvesting ...... 3...... 1 1.2.1 Resources 3 and marketing ...... 1.2.2 Processing . . . 3 Consumption...... 1.2.3 . 5..3 1.3 Other Sectors of the Economy ...... 5 1.4 Contexts ...... 5 and Decisions ...... 1.5 Information ...... " 6 1.5.1 Post-harvest losses and underutilized resources ...... 6 1.5.2 Resource assessment ...... 9 1.5.3 Infrastructure 9 in effort (or catch) ...... 1.5.4 Reductions 10 Increases in effort (or catch) ...... 1.5.5 ..10 1.6 Summary ...... 11 II. THE IMPORTANCE OF SOCIOCULTURAL INFORMATION ..11 2.1 Introduction ...... 11 Relationship Between Technology and Social Organizatin 2.2 ...... 13 Acceptance of Change and the Success of Development Projects 2.3 . ... 14 Change, Purchasing Power and Social Stratification 2.4 Technological .... .15 2.5 Summary ...... 17 III. INFORMATION REQUIREMENTS ...... 17 3.1 Introduction ...... Information ...... 18 3.2 Biological/Ecological .. 18 Fishery resources ...... 3.2.1 .18 3.2.2 Unit stocks ...... 18 3 .2.? Background information assessment ...... 19 3.2.4 Resource .19 3.24.1 Preliminary assessments ...... 20 3.2.4.2 Quantitative resource assessment ...... 21 3.2.4.3 Selection of an appropriate yi ,- model...... 22 3.2.4.4 Resource surveys ...... 23 3.2.4.5 Empirical yield models ...... 23 3.2.4.6 Multispecies fisheries ...... 23 Information ...... 3.3 Sociocultural .23 Introduction ...... 3.3.1 .23 General background information ...... 3.3.2 .24 structure of the occupation ...... 3.3.3 Social .24 occupational preference and training ...... 3.3.4 Innovation, .25 Information ...... 3.4 Economic .25 3.4.1 Introduction ...... 25 3.4.2 The bioeconomic state of harvesting ..... Open accet...... 25 3.4.2.1 34 3.4.2.2 Resource rents and owners ...... 35 3.4.2.3 Other options ...... 37 3.4.2.4 Many species and mixed gear ...... 39 3.4.3 The delivery system ...... Previous Page Blank 3.4.3.1 The industrial catch ...... 41 3.4.3.2 Consumer demand ...... 41 3.4.3.3 Supply ...... 43 3.4.3.4 Bottlenecks, power and losses ...... 43 3.4.4 Information requirements ...... 46 3.4.4.1 Transactions ...... 46 3.4.4.2 Productive processes ...... 46

IV. DATA REQUIREMENTS ...... 49 4.1 Biological/Ecological Data ...... 49 4.1.1 Habitat/ecosystem data ...... 49 4,1.1.1 Biological/ecological inventory ...... 49 4.1.1.2 Extent of fishing grounds ...... 49 4.1.1.3 Production rates and ecosystem dynamics ...... 49 4.1.1.4 Hydrographic data ...... 50 4.1.1.5 Other physical/chemical data ...... 50 4.1.2 Stock assessment data ...... 50 4.1.2.1 Introduction ...... 50 4.1.2.2 Catch and effort data ...... 50 4.1.2.3 Vital statistics ...... 52 4.1.2.4 Exploratory fiehing surveys ...... 53 4.2 Sociocultural Data ...... 54 4.2.1 Introduction ...... 54 4.2.2 General background information ...... 54 4.2.2.1 Identification of groups ...... 54 4.2.2.2 Degrees of intergroup tenaion ...... 54 4.2.2.3 Numbers and locationui of potential project participants . 54 4.2.2.4 Traditional communication channels ...... 54 4.2.2.5 Level of community dcvelopment...... 54 4.2.2.6 Occupational structure ...... 55 4.2.2.7 Numbers of small-scale fishermen ...... 55 4.2.2.8 Availability of alternative occupations for fishermen . . 55 4.2.2.9 Temporal distiTi.ution of fishing effort ...... 55 4.2.2.10 Local knowledge about fishing and fish ...... r 4.2.3 Social structure of the occupation ...... 56 4.2.3.1 Fishing gear types and ownership patterns ...... 55 4.2.3.2 Crew size and social composition ...... 55 4.2.3.3 Criteria for crew selection ...... 55 4.2.3.4 Degree of occupational mobility ...... 55 4.2.3.5 Interaction between crew members and ...... 56 between owner and crew 4.2.3.6 Degree of on and off vessel ...... 56 occupational specialization 4.2.3.7 Relationship of fishing groups with other ...... 56 social groups in society 4.2.3.8 Local distribution of wealth and power ...... 56 4.2.4 Innovation data needs...... 56 4.2.4.1 History of innovative behavior ...... 56 4.2.4.2 Attitudes towards risk, change and investment ...... 56 4.2.5 Occupational preference and training data needs ...... 57 4.2.5.1 Attitudes towards fishing by non-fishermen ...... 57 4.2.5.2 Attitudes of fishermen towards fishing and ...... 57 alternative occupations

V 57 4.2.5.3 Attitudes towards income changes ...... 57 4.2.5.4 Perceived innovation complexity ...... 57 4.2.5.5 Perceived innovation compatibility ...... 57 4.2.5.6 Perceived relative advantage of innovation ...... 57 4.2.5.7 Perceived trialability of innovation ...... 58 4.2.5.8 Perceived observability of innovation ...... 58 4.2.5.9 Irdividual attributes associated with innovativeness ...... 58 Economic Di ...... 4.3 .58 4.3.1 Orientation ...... 59 4.3.2 "Economic" data ...... 61 4.3.3 Harvest sector activity ...... 62 4.3.4 Variable definitions ...... 62 4.3.4.1 Identifiers ...... 62 4.3.4.2 Revenue and catch ...... 65 4.3.4.3 Costs in general ...... 65 4.3.4.4 Variable costs-general ...... 66 4.3.4.5 Variable costs independent of catch ...... 67 4.3.4.6 Costs which vary with catch alone ...... 67 4.3.4.7 Labor costs ...... 69 4.3.4.8 Other considerations ...... 70 4.3.4.9 Physicpl production and effort data ...... 4.3.4.10 Maintenance and repair costs ...... 71 ...... 73 4.3.5 Fixed costs ...... 74 4.3.6 Industrial fishery data ...... 77 V. DATA COLLECTION METHODS ...... 77 5.1 Introduction ...... 77 5.1.1 Wrilten records ...... 77 5.1.2 Key informant interviews ...... 78 5.1.3 Census surveys ...... 78 5.1.4 Sales receipts ...... 79 5.1.5 Sample surveys ...... 80 5.1.6 Research activities ...... 81 5.1.7 The interviews ...... 81 Biological/Ecological Data Collection Methods ...... 5.2 .81 5.2.1 Habitat/ecosystem data ...... 81 5.2.1.1 Biological/ecological inventory ...... 82 5.2.1.2 Extent of fishing grounds ...... 82 5.2.1.3 Pred,.ction rates and ecosystem dynamics ...... 82 5.2.1.4 Hydrographic data ...... 82 5.2.1.5 Other physical/chemical data ...... 83 Stock assessment data ...... 5.2.2 .. 83 5.2.2.1 Catch and effort data ...... 88 5.2.2.2 Vital statistics ...... 91 5.3 Sociocultural Data Collection Methods ...... 91 5.3.1 Introduction ...... 92 5.3.2 Examination of available records ...... 92 5.3.3 Interviews with key informants ...... 92 Data from sample surveys ...... 5.3.4 .93 Economic Data Collection ...... 5.4 .93 5.4.1 Introduction ...... 94 5.4.2 Kinds of variables and sources of variation ....

vi 5.4.3 Sampling and aggregating ...... 94 5.4.4 Variation in numbers of stratified units of activity ...... 95 5.4.5 Sampling and aggregating with sales receipts ...... 96 5.4.6 Fixed costs ...... 98 VI. DATA COLLECTION STRATEGIES ...... 99 6.1 Introduction ...... 99 6.2 Integrating Data Collection Efforts ...... 99 6.3 Two Alternative Data Collection Strategies ...... 100 6.3.1 Strategy based cn static frame surveys ...... 100 6.3.2 Strategy based on continuous frame surveys...... 101 APPENDIX A. Artisanal Fishery Backgrou ' Information ...... 107 APPENDIX B. Detailed Sales Receipt ...... 110 APPENDIX C. Combined Post-Trip Interview and Catch/Effort Survey...... 111 APPENDIX D. Village Based Interview: Fixed Cost Questions ...... 115 BIBLIOGRAPHY ...... 119

vii Chapter I

THE FISHERY AND THE ECONOMY: INFORMATION FOR DECISIONS

latitudes. Small-scale FISHERY SYSTEM ies exist in tropical 1.1 THE fishing is conducted in three types of tropi­ or network of interre- cal marine environment: 1)the coastal shelf A fishery is a system and islands, 2) estu­ lated activities which includes the harvest- platfom.Ls of continents and consumer aries. anl 3) coral reefs. Reefs and estuaries ing, processing, marketing domains of the demand for fish (Figure 1). A fishery oper- are usually the exclusive socioeconomic and polit- small-scale fishermen; competition with ates within certain is more com­ ical contexts and interacts with other large-scale industrial fisheries sectors'of the economy. It functions because mon in Ehallow coastal waters. the participants, btriv-ng to satisfy their com­ such goals as In general, tropical ecosystems are basic needs and to achieve of species. In addi­ self-respect and peer esteem, posed of a large number economic gain, average size of the fish' which are cooperate and compete with each other. tion, the be con- harvested by tropical small-scale fisheries Some aspects of the system can reef ecosystems (a fisherman's deci- is often quite small. Coral trolled by individuals by a complex network of sion to fish, for example). Other aspects of ate characterized the inter.species relationships and a high rate the system (such as the size of the fleet, which is fluctuations in of biological production, most of weather, and natural the ecosystem. Tropical abundance) are beyond the control consumed within resource estuaries are characterized by highly sea­ of individuals and sometimes beyond the salin­ working sonal river flow, seasonal changes in control of all the participants and much more constant Collective action is required when ity distributions together. temperatures than are found in temperate individual actions fail to produce desirable as important organizations, for zone estuaries; they also serve results. Fishermen's areas for many coastal species. example, can buy supplies in bulk to reduce nursery a well- Considerable organic matter is derived costs while governments have especially man­ role in providing public goods. from bordering vegetation, established groves. All ofthese ecosyatems are suscepti­ Governments also invest in facilities such such ap. ad bridges which reduce ble to environmental perturbations as wharves, roads caused by contaminatioii, high the cost of operating the fishery. Finally, those in manag- temperatures brought about by deforesta­ governments often act for society and domestic uses of property fishery resources tion, dams, industrial ing common water, the physical alteration of ceastal which have no owners. habitats and fishing. SMALL-SCALE FISHERYhaitsndfhng 1.2 THE by a Small-scale fisheries are characterized vessel types. Fishing 1.2.1 Resources and Harvesting variety of gear and techniques are generally labor intensive; This manual focuses on small-scale fisher- the types of gear used are diverse and rela­ small­ ies conducted in coastal marine waters of tively inexpensive to operate. The developing countries. Most of these fisher- scale fisherman and his family are usually

bony fishes, crustaceans or mollusks. inLludes any type of animal which is harvested, such as sharks, ' The term "fish" as clams, shrimpand lobster and do not harvest many Many small-scale fisheries depend primarily on invertebrates such bony fish at aU. GOVERN4MENT THE SMALL-SCALE FISHERY SYSTEM

RESOURCE & HABITAT OTHER SECTORS OF THE ECONOMY HARVESTING INDUSTRIAL FISHERIES

AGRICULTURE PROCESSING

INDUSTRY CONSUMPTION

SOCIETY

Figure 1.Diagram showing elements of the small-scale fishery system and its relationship with government and other sectors of the economy. among the poorest of the poor. Their income courage fishermen from selling to competi­ from fishing is extremely variable and, in tors. These arrangements are important for no general, they have little control over prices small-scale fishermen who have few or which they are paid for fish. alternative sources of credit. The relation­ ships between buyer and seller are fre­ Regardless of how efficiently the fishery quently based on extended family ties or functions, its potential contribution to other sociocultural groupings within the society ii.ultimately limited by the size and community; they also affect the efficiency of productivity of the resources which are the procedures which are used to market harvested. Fishery resources are renewable fish. produce surplus biomass which can be and can increase the harvested (Figure 2). If harvest is excessive, Processing maintains and replace value of fish: it allows fish to be shipped birth and growth processes may not a biomass lost to natural mor- farther, stored longer and converted into the quantity of types of process­ tality and harvesting. A common objective more desirable form. The is to maintain a pop- ing used in small-scale fisheries are rela­ of fishery management smoking and ulation size which produces the maximum tively simple (drying, salting, rate of population growth and therefore the icing, for example). yield to the fish ,ry maximum sustainable intervention inmarketing and in Figure 2). Government (MSY processing is stimulated by the desire to sys­ Most small-scale fishery resources are free improve the efficiency with which the who desires to do tem operates, thereby increasing the quan­ to be harvested by anyone It so. This open acress nature of the fishery tity and quality of fish which is available. biological overfishing also provides credit and protects public frequently leads to for (beyond MSY) and more frequently to eco- health by setting minimum standards (beyond MEY, Figure 2) fish as it moves through the system. Credit nomic overfishing lending to a point where the total cost of fishing is may also be provided by private equal to the total revenue gained from fish- institutions or by fishenmen's organizations. ing. While MEY (maximum economic yieii) may in rare instances be to the right of 1.2.3 Consumption MSY, the maximum economic benefit to the In many eountries fish products provide an nation from the fishery is usually achieved important source of animal protein. The to the left ofMSY. 2 This matter is discussed price paid for fish products in relation to in greater detail in Chapter 1II. other meats varies widely from country to is relied upon and Marketing country. Most frequently fish 1.2.2 Processing more by the poor as a protein source than it many Small-scale fisheries are characterized by a is by the more wealthy. Fish, like systems, ranging from other foods, may be rejected by the consu­ variety of marketing prevent people individuals who purchase fish as soon as it mer because cultural taboos it in the streets of the from eating it or because it doesn't "look is landed and sell result in sig­ village to fairly sophisticated marketing god." Consumer rejection can networks involving a number of middle- nificant waste. and the men, some kind of processing, frequently been of fish to distant markets. Governments have transportation instimulating consumer demand Fish is purchased and sold at each stage in involved to the eventual for fish products in an attempt to increase the process, thus adding of l-cotein. As we price paid by the consumer. Each partici- per capita consumption certain finan- will see later, increased consumer demand pant in the procees assumes improv, the order to earn an income, for fish does not necessarily cial risks in fishermen's welfare. usually character­ Small-scale fisheries are OF THE ECON- ized by considerable variations in supply, a 1.3 OTHER SECTORS factor which results in fluctuations in OMY income. Thus, in order to obtain a more reli­ able supply of fish, dealers often provide The relationship of other productive sectors loans and other financial incentives to dis- of the economy to the small-scale fishery

to the relates yield to the size of the resource and to the amount of fishing effort which lies 2Apoint on the curve which to the left of MSY denotes less right of MSY (Figure 2) denotes more fishing effort and a smaller population size; a point effort and a larger population size.

3 MSY TC

MEY YeTC=TR

(or Revenue)

Fishing Effort

Figure 2. Yield-effort curve for an exploited fishery resource showing how equilibrium yield changes as fishing effort increases (and population size decreases). This model is based on the premise that equilibrium yield is equivalent to the 'ate of increase in population size and that maximum sustainable yield (MSV) isreached atone-halfthemaximum amountof effort (andhalf the maximum population size). Ifyield is multiplied times price, the curve becomes atotal revenue (TR)curve. Furthermore, iftotal costs (C) of effort increase pronortionately with effort, a point is reached where TC =TP. Maximum economic yield (MEY) is achieved when total revenue exceeds total cost by the maximum amount. also be considered. This relationship probably have to be imported, thus influenc­ must balance of is defined by how other sectors affect 1) ing in a small way the country's inputs to the fishery system, 2) the opera- payments. tion of the system, and 3) outputs from the system. The existence of an industrial fishery, for example, can have a dramatic 1.4 CONTEXTS effect on a small-scale fishery. On the posi- The small-scale fishery system operates tive side, the industrial fishery may provide within economic, physical, sociocultural, the impetus for the development of wharves, legal, institutional and political contexts. In roads and marketing systems. It may also general, the contexts determine how the sys­ provide an export outlet for small-scale ter and its participants operate. They fishery production. On the negative aide, define what is allowable, acceptable and there will be conflict between the industrial desirable. The contexts set limits on how the fishery and the small-scale fishery if they system and its participants will respond to both harvest the same resources, fish in the changes - expected changes such is price same areas, or if the industrial fishery fluctuations and the gradual introduction of increases the mortality rate of species new fishing practices which take place at which are exploited by small-scale fisher- their own pace and accelerated changes men.' Frequently the small percentage of (such as the introduction of new fishing by-catch which is sold by industrial fleets is gear or decisions to increase or reduce effort) sufficient to depress prices paid to the small- which result from deliberate interventions. scale fisherman. Attempts to intervene at some point in the Activities in the agricultural sector may system without adequate consideration of affect.the small-scale fishery for many rea- the possible effects on the entire system and sons: many, if not most, fishing families its various contexts incur considerable risk also raise crops and livestock; the agricultu- of failure. Changes which are unacceptable ral sector may dominate the regional distri- in the economic, sociocultural, institutional bution and marketing network and thereby or political contexts can result in failures as define the means available for expanding certain as changes which are, for example, the distribution and marketing of fish; the technical failures. These determinants of use of agricultural pesticides and herbicides failure are the lessons of development least can threaten the survival of fish and make learned. them unsafe for human consumption. On the other hand, fishing and agriculture can In the broad political context the extent of complement each other. For example, by- the government's intervention in the products of fish processing can be used for fishery is clearly related to the fishery's sig­ fertilizers and animal feed. One ofthe most nificance in national planning. If fishery important relationships has to do with the development is perceived as being of minor supply of labor. The number of people wbo significance or in conflict with other poli­ fish and the amount of fishing they do is cies concerning the use of the ocean and its closely related to the returns they can expect shoreline, important support structures in fishing relative to their returns in agricul- such vs government fisheries offices, exten­ ture. The multiple relationships which exist sion services, and regulatory agencies may between these two sectors clearly indicate be overlooked. On the other hand, it is clear that changes in one sector cannot be consi- that the fishery administrator is responsi­ dered without taking into account the poten- ble for generating information which will on the other. ensure thatthe fishery is properly evaluated tial effects in the process of establishing national related to riorities. operation ofthe fishery is also The services provided the credit and marketing by the commercial sector. If fishing is viewed as a risky investment as compared 1.5 INFORMATION AND DECISIONS to other investment opportunities, financ­ subsidy. Ifthe The last section of this chapter describes ing may require government aimed sector cannot provide the tech- how a hypothetical decision process industrial consumption of nologies required by the fishery, they will at increasing per capita

3Increased mortality is caused, for example, when a shrimp trawler discards fish which are too small to have any market by the small-scale fleet. value. These fish are dead when they are discarded and will never be available for capture

5 fresh fish might be carried out.4 The fishery acceptable to the consumer, or they may be in question is a very simplified one - dev .'i more subtle changes such as the recruiting of most real world complications. The deci- of crew members who are willing to spend sion process is examined solely to demon- longer periods of time at sea. strate its multi-disciplinary nature and the wide variety of information wh eh is 1.5.2 Resource Assessment This exercise is not intended to required. losses represent a plan for action. If production is low and post-harvest are minimal, it becomes important to deter­ In Figures 3 and 4 it is assumed that ade- mine whether or not there are sufficient fish quate analyses have concluded: 1)there is a resources available to support increased need to increase the supply of animal pro- exploitation. At this stage in the decision­ tein; 2) the fishery appears to present a rea- making process, a simple preliminary eval­ sonable alternative for increasing this uation of the resources can be based on supply; and 3) sufficient fresh fish is not interviews with individualS who are famil­ available at current prices to meet projected iar with the changes which have taken demand. It is at this point interventions place in the fishery (specis, quantities and aimed at increasing the supply of fresh fish sizes of fish; types and number of vessels, should be evaluated. 5 gear or fishermen) and on historical records of catch and effort, if they are available. A 1.5.1 Post-Harvest Losses and Unde- preliminary evaluation only permits quali­ rutilized Resources tative judgments of the degree of over- or underexploitation. A more thorough assess­ One of the most cost effective ways of ment should define more exactly the degree increasing the supply of fish is to utilize fish of over-or underexploitation and predict the which is already caught but never reaches expected effects of certain management the consumer. Post-harvest losses are strategies on resource abundance and yield. caused by the failure to use ice or the improper use of ice during harvesting, and If a preliminary evaluation suggests that by poorhandling and insufficient storageor resources have been overexploited, then distribution facilities once the fish is some intervention may be necessary to landed. In addition to reducing these losses, reduce catch and/or effort. At the same another effective method for increasing time, a more thorough assessment should be supply would be to reduce the quantity of initiated in order to estimate how much fish that is discarded at sea during commer- increased production can be expected as a cial fishing operations. Decisions to utilize result of certain management strategies, by-catch require economic evaluations of and a mechanism for collecting the neces­ processing and marketing possibilities. sary data should be established as soon as Fish which have no value in the fresh fish possible. Also, a search for new, unexploited market might require some form of process- resources should be considered. ing and market promotion. If currently harvested resources appear to Other important sources of additional fish be underexploited or ifnew resources which protein are resources w'hich are not pres- are not being utilized can be identified, ently being harvested or which are har- attempts to increase pcoduction can pro­ vested, but are not directly utilized for human ceed, but should proceed ,dowly while data consumption. Significant increases in fresh for a more detailed stock evluation are col­ fish production can be realized by making lected. Development of new resources which changes which improve the capture and/or have not been assessed should proceed in marketing of underutilized resources. These well-defined stages so that changes in bio­ changes may be fbvious ones such as the logical and economic parameters can be use of more effective fishing gear or a type of evaluated as production increases. A simple processing which makes the product more monitoring of changes in catch per unit

4A will be discussed further in Chapter III, this goal is frequently inconflict with the goal of assisting fishermen and their families.

5There are, of course, other paths that could lead to a decision to stimulate increased production. For example, foreign exchange can be generated from a resource that commands high international prices. Somesmall-scale fishery resources (snapper or spiny lobster, for example) ccmmand high international prices; increased production and export of these resources generates foreign exchbage.

6 EVIOEU1TAL CAUSES. OlIJCtIVE: LOW FIODUCTIVITIY, UKMM SUPLYe.g. LI]ITED FISHING GMUNDS Of SL not support willIncreased production t

to increase production CS/EVTAppears that Attempts A OFI FISHSRY OF wSScS CustRESOURCE VERSUS 3.IALhyoteharves e rs reIncAUTshould be delayed

EAUTOOFwill support EXCESSIVE HARVEST increased ishepryc 10oREiURC production wilre duter ESORCE e hrvesed EDUCE EFFORT8~snn beneits morer fis cand~ledrsuc

ASSESSASSES5

Infrastructure DMIAND FOR FIS needINRAEFOT Projected ~do"snot exceedssupplyneeds Iprovin

pie Projected des"n exceeds supply at current prices 7atIFATUUR ote: Attempts to increas production are i*Lg-RZ-ASESS before a more detailed resource &""=sOut CiOMRFISHN LaiatFOcompleted should proceed slowly IHi ASSESS MlAGNITUDE LOSSES 1

LossesUTILIZATION OF Israel ASTEDFISH

when attempt­ Figure 3. A hypothetical series of actions and decisions which must be considered igto expand the fishery's role in increasing protein supply. BtoSOCCS(iC COSTI3MMT Moployimst ANALYSI 0 ALTRlNMIVE Presure on resource itu Ownership patterns Availability ot napital Availability of skilled labor Tuaining coats

APPOPIATE-

ASSESAVAILASUSSLATTIT DJETST~TI,

ASSESSo AASAILAVABILITYT

SkilledASESVALBIT labor Is not OT CAPITAL available labor Is availableq

RA Suffcent Sufficient capital Ig not capihnl Is

available

84

ASSESS M4ECHANISM

Machanisa dahso ntne need todootue be isprowd to be imqproved

YOM UmTSOG CUIDIT - AV&IAIIM- OR DrV11w w oMl

EUM"L CMIUX

Figur 4. A hypothetical series of actions and decisions generated by an attempt to increase production from the fishery by increasing fishing effort.

8 not support effort, for example, and in the economic sizes are small and will of different technologies over increased harvesting, the possible explana­ profitability Is resource abun­ time can provide useful information for tions should be examined. and management dance limited by environmental factors making development unfavor­ when thorough stock assessments such as poor primary productivity, decisions or limited fishing are not feasible. able climatic conditions grounds or has it been reduced by excessive Decisions to proceed with more detailed harvesting? If stocks were previously more resource assessments cannot be made abundant, but have been recently depleted Detailed analyses can require con- by intensive fishing effort, a reduction in lightly. to expenditures of time and money. effort or catch would eventually lead 7 siderable and increased production. However, even in regions where there is lit- stock recovery and exbdng fishermen seem to tle fishing are small, but were have no trouble catching as much fish as In cases where stocks detailed stock assessments never very large, the gains in production they want, of effort be considered in cases where which would follow any regulation should the social and eco­ increased fishing pressure is anticipated. would probably not pay nomic costs of management. The societal 1.5.3 Infrastructure costs of any regulatory scheme include for those which are immediately obvious, If a preliminary resource assessment sug- example, those associated with designing, gests that stocks will support increased negotiating, managing and enforcing regu­ exploitation -or that post-harvest losses can lations and those which are less obvious be reduced - then the potential for expand- such as the costs of developing alternative ing marketing, transport, processing and employment opportunities and retraining distribution services should be evaluated fishermen, or the costs of unemployment before attempts to increase production are and urban migration. made. This means examining the delivery system for obvious bottlenecks in storage, Regulations are usually intended to limit If suffi- the size at which fish are first captured transport or marketing services. 8 cient infrastructure for providing these ser- and/or to limit effort. Regulations which vices exists, then development can proceed are related to size protect the reproductive while data are being collected for a more capacity of the stocks by ensuring that thorough stock assessment. However, ifthe enough sexually mature adults remain in nat­ infrastructure needs improvement and an the population to replace fish which die of the anticipated benefits and urs"y or are harvested. Examples of this analysis size costs of the improvement is positive (indi- type of regulation include minimum cating that the benefits exceed the costs), limits, gear restrictions which affect the size development plans which include infra- of fish captured and closures in particular structure improvement can be made while locations or times of year when small fish additional resource assessment data are col- are more abundant. Effort regulations are lected." If the analysis is negative, then intended to limit the catch to a certain level. attempts to increase small-scale fishery pro- Some of the more common ones are closed duction should probably be abandoned or seasons or areas, gear restrictions, catch reformulated in terms of overall rural devel- quotas and limits on the number of boats or opment objectives, fishermen. 1.5.4 Reductions in Effort (or Catch) However, none of the management strate­ gies mentioned above are successful in pre­ If resource assessments indicate that stock venting the over-investment of labor and

improvement unless extensive underutil­ that small-scale fisheries alone would justify major infrastructure 6It is unlikely be on a large scale, however, to improve the way in ized stocks of fish are found. Infrastructure improvement need not which the fishery functions.

p-oduction until the fish which 7 The immediate result of a reduction in effort on an overexploited stock is a decline in number of offsoring. escape capture, grow larger and produce a greater

practice, regulations aimed at reducing effort two management strategies are not independent of each other. In s These eventually grow to reach a larger size. result in a greater average size since more fish escape capture and

9 capital in the fishery. Effort restrictions, for light fishing and can dramatically increase example, which are directed at limiting only catches, thereby depleting resources, over­ certain components of effort (time, location loading the marketing network and per­ or the efficiency of capture) - but not total haps reducing prices paid to the fishermen. effort - may reduce catch, but will result in Short term gains such as increased produc­ the waste of society's human and financial tion may very well be followed by long term resources. Only in the case where someone losses in resource abundance, damaged per­ acts as the owner of the resources and sonal relationships between fishermen and charges for their use is there proper eco- a less desirable income distribution. nomic exploitation. Examples of regula­ tions which simulate this type of ownership 1.6 SUMMARY are taxes on effort or catch, or marketable individual boat quotas or licenses. A combi- In sum, a hypothetical decision process onnation total of effort a size-related is often ideal.restriction9 with one ofaimed fresh at fishincreasing was examined er capita to consumptiondemonstrate its multi-disciplinary nature and the wide 1.5.5 Increases in Effort (or Catch) variety of information required. Following a brief discussion of the importance of socio­ If it is determined that the stocks can sup- cultural information in the next chapter, the port increased fishing effort because, for remainder of the guide will identify specific example, 1) the area they inhabit is larger information needs, the data from which the than previously believed, 2) human popula- information is derived, and appropriate tion pressures have not as yet led to their data collection methods. overexploitation, or 3) new stocks can be exploited if the range of the boats can be increased, then the next step is to determine by what means to increase effort. It must be realized that there can be a trade­ off between seeking the least costly method of extracting additional quantities of fish and other goals such as increasing employ­ meit and promoting rural development. For example, many small-scale fishing tech­ riques are extremely efficient in their use of both labor and capital despite the fact that fishing trips are of short duration (a day or less) and the limited range of the vessels limits most fishing to nearshore grounds. Economies of scale can often be realized with larger vessels but increased efficiency must be weighed against the possible risks of favoring certain individuals (wealthy versus poor fishermen, for example), as well as the ability of the participants tr obtain financing and the existence of credit sour­ ces - to name just a few of the important economic and social considerations. The potential impact of the proposed changes in harvesting strategy on the resources must also be examined. Figure 4 outlines some of these considerations. Caution is warranted here; there is a long history of failure associated with technolog­ ical innovations. A technological change as simple as replacing multifilament nylon gill nets with monofilament nets permits day­

9 The plight of the remaining fishermen is not necessarily any betterafter a reduction in total effort. This is discussed in Chapter III.

10 Chapter II

THE IMPORTANCE OF SOCIOCULTURAL INFORMATION

experience, developed a model which indi­ 2.1 INTRODUCTION cated that vessel size and complexity affect studies have indicated both crew size and the recruitment of crew Although numerous noted, how­ success or failure offishery develop- on the basis of skill. It should be that the equipment which projects depends largely on ever, that more efficient me-t crew size may result in sociocultural factors, they are often over- reduces necessary projects are being unemployment and increased social strati­ looked when development that rationale for biological and fication. The model also indicates planned. The the basis of information is relatively well crews are often selected on economic as kin group member­ developed and widely accepted. This is not social criteria such with respect to sociocul- ship. Small-scale fishermen work-groups, the case, however, in structure ingormation. Sociocultural informa- however, tend to be egalitarian tural many shipboard tasks tion has two important functiona in fishery due to the fact thaz to addressing require close cooperation between fisher­ development. In addition ties management issues, it men. These close interdependent development and in the economic and biological data col- between crew members often result facilitates ashore based on Identification of social groupings of fbrmation of male groups lection. prevails at sea. fishermen, their informal and formal lead- the work-group which their systems of com­ ership patterns and the model notes that the gen­ munication provide information which can Furthermore, erally low cost of small-scale fishing tech­ be used to structure effective data gathering of the the likelihood of nology, the impermanent nature systems while increasing to the destructive nature of cooperation of the fishermen. equipment (due obtaining the the sea) and the close on board cooperation which is required usually result in little 2.2 RELATIONSHIP BETWEEN and AND SOCIAL social distinction between owner TECHNOLOGY small-scale fishing groups. ORGANIZATION laborer within Nevertheless, as equipment costs increase like­ Several aspects of small-scale fishery tech- due to increased size or complexity, the result in social relationships that lihood of ownership by individual fisher­ nology the differ somewhat from those found in agrar- men decreases, thus promoting groups. It is therefore important development of social stratification and ian social of to examine the relationship between small- inequality. Additionally, increased costs fishing technology and certain capital equipment often lead to the develop­ scale as aspects of the ownership of equipment, as ment of financing specialists. Finally, well as work-group and non-work-group production increases, there is an increasing structure and the degree of social stratifica- need for distribution and processing tion. An understanding of these relation- specialists. allows one to predict potential costs ohips between these various accompanying alternative processes of The relationships and thus enables the aspects of small-scale fishing technology technological change, are illustrated in to weigh these costs against the and social organization planner are needed at all points in the benefits of introducing the technology. Figure 5. Data decision-making procees in order to predict Pollnac (1982; 1979), draw- changes that will result fiom a given tech­ In earlier papers, it is ing on available literature and f-search nological change. To predict changes,

11 YON DEGREE OF SOCIAL I I STRATIFICATION

WODEGREE OF INCRASEDDISRIBUIONOCCUPATIONAL PRODUCTION SPECIALIZATION

EQUIMENTOWNERSHiP COST [TlYEPE

TECHNOLOGICALU

DIVERSITYA INTERD)EPENDENCE VESSEL SIZE AW CREW SIZE AND COOPERATION OF CREW UNEMPLOYMENT 1CCUPAION _LIOBLI Y [SOCIAL DISTANCE BETWEEN WORKER AND OWNER

EGALITARIAN WORKGROUP

EORKGROUPS ASHORE

Figure 5. Relationships between technology and social organization (adapted from Pollnac, 1981).

12 necessary to evaluate the technology in The return to individual fishing did much to terms of its complexity, labor requirements, restore good relations, but the degree ofcom­ cost, and productivity - variables which munity organization which was originally play key roles in the relationship between based on boat crew memb :rship and the technology and social organization. traditional authority of the orang baik (whose traditional status depended on boat At this stage of the discussion it is perhaps group affiliation) was never regained. appropriate to introduce several examples Further, 'he introduction of nylon nets and which will demonstrate the interrelation- individual motorized vessels reduced the ships between technological and institu- need for a large crew; nevertheless, the tional changes and aspects of social crews were kept larger than necessary in organization contained within the model. A keeping with traditional crew structure. study conducted by Fraser (1966) among Fraser (1966) argues that the maintenance Malay fishermen of South Thailand pro- of large crews plus decreasing catches vides a good example of the impact techno- undermined the sense of pride that tradi­ logical change can have on work groups tionally characterized crews. This decrease and community social structure. in group solidarity reduced the relatively Traditionally, the Malay fishermen of high status of the steerer and, hence, his ru adn lite onMoarsland aishermenaofstatus in the community at large. Thus, a relied on oars and sail to take change in technology that was poorly Rusembilan of them to their fishing grounds. In 1956, adapted to the traditional social structure groups of boat owners and steerers (tradi- work was rejected, and the negative impact tionally a high status position in the boat on the social structure of the community crew) dominated deliberations concerning was never totally corrected. Furthermore, the best way to motorize the fleet. They labor saving technological innovations decided to introduce tow boats to take fish- were ineffective because the fishermen were ing vessels to fishing areas and bring them unwilling to use fewer than the traditional back. Groups of boats formed tow groups number of crew members. associated with a particular tow boat. This new technology immediately placed consid- A similar reluctance to change work-group erable strain on the traditional social structure was recently reported for small­ syftem. scale fishermen in Malaysia. Sabri (1977) notes that although winches were installed, First, membership in tow groups meant that thus reducing the number of fishermen individual boat crews and steerers lost their needed on a vessel, traditional crew size was previous independence with regard to locat- maintained to provide employment for ing fish and timing the return to iarket. members ofthe extended family. In another Second, after a period of poor fishing, wives area of Malaysia, however, Yap (1977) of members of the more skillful boat'crews reports that improved technology resulted they were subsidizing leas suc- in a reduction in crew size and significant realized that oe unemployment ar,,ong fishermen with no cessfultowgru' crews oasince ac.Faeshares were based16 (in the tow group's total catch. Fraser (1966) fptes alternative occupations. This impoverished that this situation had broad repercus~ions class of unemployed fishermen, of course, in other areas of community life. It resilted increased the degree of social stratification in overt hostility between women, and 'ela- within the fishing community as the model tions between men became strained. the would predict. coffee shops, which were the focus for cpm­ munity decision-making groups and a 1so- 2.3 ACCEPTANCE OF CHANGE AND ciated with boat crews, manifested a THE SUCCESS OF DEVELOPMENT marked drop in attendance, reflecting the PROJECTS social strains. Attendance at coffee sh ps never fully recovered. Further, trad" ioi al In order for production to increase as the village authority figures, the orang bi ik result of introducing new technologies or (morally good man), were involved in new fishermen into the fishery, the partici­ ownership of tow boats and their operati n. pants in the fishery system must be willing Thus, the chief source of authority a id to accept change (seeFigure 4). It is obvious, means of maintaining village control wi re but often overlooked, that if the participants undermined. Finally, because the religic us refuse to cooperate, development projects leaders of the village remained aloof fr m will not succeed. Often this reluctance to the changes, their status increased. Befilre cooperate is based on rational considera­ long the strains became too great and t he tions which can be accounted for when tow boats were eliminated, designing the project if they are known

13 beforehand. It is therefore necessary to tion. Finally, among Malay fishermen assess the attitudes, beliefs, and values of increased costs of productive equipment the participants towards the proposed associated with modernization has pro­ changes. Consideration of these factors in duced a class of equipment owners. the early planning stages is an important equip­ ingredient in projecl, success. Firth(1966) has noted that although ment modernization has resulted in greater If attitudes toward change are negative, economic returns to the entire fishery, then it is essential to determine why and increasing capital costs have led to a attempt to adjust the changes in order to marked drop in the percentage of earnings satisfy the perceived needs of the people. If going to the labor force. Despite the fact that attitudes are positive, then success in the the fishermen have become, in effect, use of a new technology or the addition of employed laborers in the new system, they more fishermen depends on the availaoility are treated as participants in a common of skilled personnel - another decision point enterprise and thus not put on a regular where detailed information is needed (see wage basis. Their income is still based on a Figure 4). If there are not enough skilled share of the catch. Specifically, Firth (1966) personnel, some sort of training program has noted that among the Malay fishermen, should be considered to either retrain exist- costs are removed from the catch before ing fishermen in the use of the new technol- shares are calculated; thus, given the perio­ ogy or teach unemployed people how to fish. dic nature of production in the marine envi­ ronment, fishermen often receive next to 2.4 TECHNOLOGICAL CHANGE, nothing. He therefore reports that in 1963, PURCHASING POWER AND SOCIAL the fishermen were in a less advantageous STRATIFICATION position than when he first studied them in 1939-1940, and that the entrepreneura were Once sufficient skilled personnel are availa- much more economically powerful than ble, the only remaining obstacle which their predecessors of a generation earlier. could still prevent fishermen from utilizing a new capture technology is their ability to Even when governments are aware that the purchase the equipment. If sufficient funds high initial costs of new technologies can are not available, methods for extending increase social stratification, problems credit must be considered. Ifthe distribution often persist and increased disparity in of capital is very unequal, the introduction wealth results. For example, Alexander of new technologies could result in (1975) reports that in Sri Lanka the govern­ increased social stratification, a process ment was aware of financing problems which is often accompanied by other social associated with costly new fishing technol­ problems (Pollnac, 1976). In many cases the ogy, so they introduced a hire/purchase only individuals who can effectively take scheme. Individuals who took part were advantage of new opportunities are those selected by ballot from qualified applicants. who are already wealthy and the new tech- The individual fisherman had to provide a nolhgy only enhances their situation in rela- deposit and received a government loan, tion to others. The model presented in repayable over five years, to purchase a Figure 5 outlines the relationship between boat hull with an engine. Unforeseen prob­ technological change, equipment cost, lems developed, however. patterns, and social equipment ownership with the stratification. First, the deposit, in combination fact that the loan covered vessel and engine Once again, some examples from fishery but not gear, meant that the fishermen had development projects are in order. Epple to borrow money from private noney lend­ (1977) provides a good example of how ers. Second, the new equipment deteriorated mechanization, because of the increased faster than the old, and there was no provi­ price of capital equipment, altered patterns sion for maintenance funds. Third, loan offishing boat ownership on Grenada. Prior repayment was not related to the value of to mechanization, 90 percent of the fisher- the catch -it was a fixed monthly payment; men owned their own boats. Following thus, during periods of low production the mechanization this figure dropped to 25 per- payment could exceed income. Neverthe­ cent. Sabella (1974) also noted that as Peru- less, production increased, so the govern­ vian small-scale fishermen began to depend ment viewed the project as a success and on expensive, highly specialized equipment, invested more funds in it. The total income their formally egalitarian community to the fishing village increased, but prob­ began to manifest signs of social stratifica- lems began to appear.

14 of fishermen increased technology was to be introduced to indigen­ Since the number crew­ little during the years after the new loan ous fishermen using a plan wherein scheme was first introduced, increased pop- men would collectively own the equipment. in greater unemploy- Traditionally, crewmen were bound to a ulation size resulted "loan" ment. New boats were introduced, but they boat by an interest-free permanent available only to provided by the boat owner. The boat owner rightfully were made a similar experienced fishermen. Important for our was bound to a money lender by is the fact that inexpe- arrangement. According to Emmerson, the discussion, however, the relation­ rienced new fishermen were recruited only participants did not perceive of relatives; therefore, few ship as exploitative -it was one of reciprocal from the pool in, and viewed opportunities existed for those not related to obligations, freely engaged acquire the expe- as being fair. When this traditional system the boat owning elite to of the rience necessary to qualify for operating a was threatened by the introduction of elite in the community new equipment, the fishermen destroyed boat. The number project increased substantially with the bulk oi the the equipment and assaulted a population being reduced to the poverty administrator. suggested that since level. Alexander (1975) with the extension of credit the elite have political power and control Other problems most favorable occupa- can be related to the fishermen's perception recruitment to the source of the loan. For tions, the degree of social stratification in of the immediate in example, in one fishery development project Sri Lanka will become even more marked the in social stratification in Malaysia, the source of credit for the future. Increases government sponsored have been attributed to similar factors in fishermen was a where costly innova- institution (a cooperative). Many of the other communities reasoned that since were introduced (Norr, 1972). fishermen in the region tions the function of the government was to help of new, relatively therm, the loans were like charity and did not Clearly, the introduction 1967). As a costly technologies can result in income dis- have to be repaid (Narkswasdi, increases in social result, the loans and the equipment that parties and associated as gifts and stratification. If such a situat;ion is deemed was provided were considered then technique, to extend the project encounteret serious difficulties. undesirable, determination of credit to poor fishermen should be inveoti- It should be clear that the (such as devel- locally appropriate structures for extending gated. If local organizations for project op.nent banks and fishermen's credit to fishermen is crucial cooperatives) for credit extension exist, then success. they should be used. If not, locally appro­ priate solutions should be developed, ideally 2.5.SUMMARY using traditional organizations if they exist In sum, the arguments presented in this drawn from (Siebel & Massing, 1974). chapter along with examples related actual fishery development projects should The distribution of wealth is closely informa­ of power in a community. make it obvious that sociocultural to the distribution critical role in developing a In addition, the process by which reciprocal tion can play a of business and fishery. Techniques for obtaining this infor­ obligations form the basis elsewhere in this social transactions may be affected. Unless mation are discussed these aspects of the social structure are guide. investigated in advance, the credit systems which are designed may be inadequate and it may notbe possible to anticipate potential problems. In some communities, attempts by development agencies to introduce costly fishing technology in a manner which would possibly reduce the potential for increased social stratification by avoiding traditional equipment owners and money lenders have failed dme to the fact that fishermen viewed the ti aditional patron­ client relationship as legitimate and the government's planned intervention as ile­ gitimate. For example, Emmerson (1975) describes a development program in Indonesia where a more complex, expensive

15 Chapter III

INFORMATION REQUIREMENTS

uses in the decision-making 3.1 INTRODUCTION has multiple pro- process. and interpretation ofdata The analysis a Data are collected and The purpose of this chapter is to describe vide information. economic and a series of steps and in the pro- minimum set of biological, compiled in which should be cesg become more and more refined until sociocultural information analysis. For available in any developing country where they form the elements of an des­ of the quantity of fish small-scale fisheries are practiced. This example, estimates to include all the by an individual vessel using a crip-ion is iot intended captured information which might particular gear type, when combined with possible types of by all vessels be important in any particular situation: similar data from trips made it type during a given such a task would be impossible. Nor is which use the same gear types of are transformed into annual intended to recommend certain time interval, more important than catch statistics. A time series of annual information as being selection of necessary infor­ catch data is combined with fishing effort others since the years and is ana- mation depends on management or develop­ data for corresponding which apply to provide an estimate of the maxi- ment priorities and objectives lyzed situation. mum quantity (weight) of fish which can be in any particular every year which will not harvested minimum set of information can be endanger the ability ol the resource to This natural defined on the basis of its usefulness in eval­ replace losses caused by fishing and in the etc.). The uating the feasibility of intervening mortality (predation, disease, impact of man­ which are required fishery and predicting the kinds of information efforts. Interven­ of analysis which are agement and development depend on the types at any point in the nature and quantity of the tions may be made performed, the but we are concerned only data and the management fishery system, available interventions which affect the resource and/or development objectives which are with and harvesting activities and the informa­ being pursued. tion which is needed to evaluate them. The the impact ofvar­ Decisions concerning what types of infor- emphasis is on evaluating ultimately iotas proposed changes before they occur, mation should be obtained to point out on the resources which are available not after. It is also important depend information which is des­ for collecting, compiling and analyzing that most of the costs and benefits cribed in this chapter is generated from data data and the anticipated resource and data collection procedures. Pre- which are collected from the of different of the fishery. The nature dicting the usefulness of different types of harvesting sector can be difficult. and usefulness of economic information information, however, fishery are dis­ a good strategy is not to rely on a from other sectors of the Thus, 3.4. single type of information, but rather to cussed in section different types of informa­ compile several are needed to satisfy this min­ at the same time, taking advantage of Data which tion set ofinformation needs are described alternative data sources and collection imum reliable in Chapter IV and relevant data collection procedures. Later, less useful or for effec­ and data col- methods in Chapter V. A strategy information can be eliminated data collection is can be reduced. Another good tive multidisciplinary lection efforts in Chapter VI. Analytical proce- strategy is to compile information which presented

17 Previous Page Blank dures are not described in detail at any point make them equally vulnerable to exploita­ in this gi ide afthough particular analyses tion by the same type of gear, which produce useful information are men­ tioned. As has been pointed out already in 3.2.2 Unit Stocks Chapter I, much of the information which is The ideal management unit is the unit required is multidisciplinary irn nature. stock, a term applied to a resource which is Nevertheless, the discussions of informa- exploited in a particular geographic loca­ tion and data needs and data collection tion and whose individual members methods which follow in this chapter and in respond similarly to fishing pressure. Thus, Chapters IV and V are organized according a group of fish with similar natural mortal­ to biological, sociocultural and economic ity, birth and grewth rates will not be issues. depleted at the same rate as another stock with different characteristics which is Some of the terminology which is used in exposed to the same amount of exploitation. technical A single species distributed over a fairly andthe followingmay be unfamiliardiscussions tois fairlymany readers. wide geographic arca may therefore be unit stocks. By the same Nevertheless, in order to describe the infor- made up of several toke a unit stock o than mation and datalatesoltios whicht smal-salefiserytoken, are needed to formu- a unit stock may include more than late solutions to small-scale fishery one species. In practice, unit stocks are diffi­ development and management problems, cult to identify on purely biological grounds, we feel this kind of language is necessary. and managers are forced to &fine them on Whenever possible, specific terms have the basis of available information. As more been defined and references to additional information is obtained, unit stock defini­ information have been included. tions may be revised. Unit stock identifica­ impractical in tropical 3.2 BIOLOGICAL/ECOLOGICAL tion is particularly small-scale fisheries since so many species INFORMATION are frequently exploited in the same areas 3.2.1 Fishery Resources with the same gears and since so little is known about their biology. in addition, The structure and operation of any fishery some species (coral reef fish, for example) is dependent on the resources which are are characterized by extremely patchy spa­ harvested. Fishery resources are defined tial distributions. according to thp species and sizes of fish which are captured, the gear which is used 3.2.3 Background Information and the processing and marketing practices which are followed in making fish available General biological information which is to the consumer. A pelagic purse-seine needed in order to determine what resources fishery, for example, harvests small school- are exploited includes a knowledge of the ing fish at or near the surface which may be species, sizes and relative quantities of fish canned for human consumption or reduced which are harvested by each fishery. Eco­ to fish meal or oil, while a bottom longline logical information should include some fishery harvests demersal (bottom) fish basic understanding of the physical habitat which are usually marketed as fresh or fro- and the ecosystem which supports the zen whole fish or fillets. Small-scale fisher- resource (i.e. the environment in which it ies generally produce a great variety of fish exists and other organisms upon which it in relatively small quantities which are usu- depends for food, which act as pr2dators or ally marketed as fresh fish. A resource may which compete for the same prey), and an be defined to include a number of species evaluation of the sources and extent of pri­ which are harvested by the same gear, or mary production which provide the organic the term may be applied to a single species. matter necessary to sustain population Ultimately, however, the biologist is con- growth. The potential maximum yield cerned with single species populations or which can be expected from any resource with populations made up of several species and its vulnerability to fishing pressure are which share certain common biological largely functions of its role in the ecosystem and/or ecological characteristics which and its life history characteristics.'

I Energy is dissipated at each successive stage in the ecosystem so that smaller herbivorous fish which feed directly on plant material, for example, produce larger populations than larger carnivorous fish which do not have as much food available to them. Furthermore, species which invest more energy in growth than reproduction tend to produce popula­ tions with large numbers of small, short lived individuals which should be harvested in bulk at relatively young ages. Species which invest more energy in reproduction than growth tend to produce populations with a fewer numberof large, long lived individuals which should he harvested selectively at relatively older ages. Species which are adapted for slow growth, lat-ge individual size, late maturity and greater longevity are generally more sensitive to fishing pressure.

18 assess­ 3.2.4 Resource Assessment which is required for a preliminary ment may also be useful in the initial stages Once this background information is avail- of a more detailed assessment. able, some kind of resource assessment can should reveal Quantitative assessments - if successful ­ be performed. Assessments are or or not the resource(s) is (are) over- indicate the degree to which resources whether and cart be used to exploited, i.e. has fishing caused the mortal- are not overexploited to exceed predict gains in production which can be ity rate of the exploited population management of growth and recruitment expected as a result of certain the natural rates assessments are that the maximum rate of population strategies. Quantitative so and time consuming. The data growth (and therefore the maximum expensive no longer being required for analysis are often not available sustainable yield or MSY) is may be size of the spawning and even when they are, the results achieved, or has the therefore, less reduced so much by fishing inconclusive. In many cases, population been assessments must be that the recruitment of young fish to the rigorous preliminary has been severely relied upon for determining whether or not exploitable population large to sup­ reduced? 2 Resources are usually reduced the resources are sufficiently below the size which produces MSY when port an expanded fishery. heavy fishing pressure is combined with sustainable Ideally, resources should be evaluated poor recruitment. Maximum In practice, theoretically attained at half the before they are overexploited. yield is the effects of increased exploi­ unexploited population size (see Figure 2). concern over tation is seldom ey:pressed until catches is a function ofthe number and begin to decline. In these cases, sufficient Recruitment for an assessment agi of spawning adults in the population information necessary conditions which affect may not be collected in time to prevent over­ and environmental where of eggs, larvae and juveniles, exploitation. Even in situations the survival and effort extreme variability in the envir- assessment data such as catch Given the number of factors which affect recruitment, statistics are available for a onmental and interpretation of a stock which has been heavily exploited years, the analysis a period of trends in the data is much easier when the may suddenly "collapse" when popula­ corresponds with a early years of the fishery - when poor egg production - are equally repre­ of unfavorable environmental condiv tion size was large period later years after the resource which reduce the survival of eggs, lar- sented with the tions depleted. vae and juvenile fish. was fishing effsrt frequently take the resource is overexploited, a reduction Increases in If absence of any resource assess­ in effort or regulations which impose limits place in the captured can ment, either as a result of the natural devel­ on the quantity or sizes offish a production. If the opment of the fishery or following result in increased of more efficieit resource is underexploited, development of deliberate introduction in effort) is a fishing gear by a development agen .y. The the fishery (i.e. an increase ­ alternative. Any manage- failure to evaluate exploited resources more feasible major develop­ ment strategy should ensure that the popu- even qualitatively -- before at a large enough size ment projects are implemented can have lation is maintained in sudden "recruitment fail- serious consequences: major increases to protect against of time can i.e. at more than one half the unexplo- fishing effort over a short period ure," stocks which are ited population size. lead to the collapse offishing pressure. ofa already under extreme may take the form Resource assessment qualitative preliminary assessment which 3.2.4.1 Preliminary Assessments simply indicates whether or not ovcrfishing Preliminary assessments are very simple is a problem or a more sophisticated quan- the a great exercises which are designed to answer titative assessment which requires more heavily Regardless ofhow it questions "is the resource deal more information. now than it has been in the past" defined, resource assessment is actually a exploited is "will additional fishing pressure most continuous process since the information and is 2These two types of overfishing are referred to as growth and recruitment overfishing, respectively. Recruitment by the gear in use. In defined as the number of young fish which survive to reach a size where they can be harvested harvesting results in the capture of so many immature fish (fish which never have an recruitment overfishing, excessive produces the maximum opportunity to reproduce) that the rate of population increase is reduced to below the size which sustainable yield.

19 likely lead to reduced or increased produc- known primary pzoduction rates and tion?" Answers to these questions must assumed rates of energy transfer from pri­ come from an examination of historical mary producers to herbivores or that the changes in landings (and effort, ifsuch data catch of demersal fish from a kno%.n area of are available) and information concerning bottom is much more or much less than can the species and sizes of fish which are har- be expected based on catch per unit area vested by particular gear types. Often, such estimates from other similar fisheries in information must be based on the recollec- similar habitats. tions of fishermen or other people who are knowledgeable with the fishery. Some- Indications of extreme resource depletion times, historical catch and effort statistics are much easier to detect. There is usually are available, an obvious increase in the number offisher­ men, boats or gear which are active in the In situations where resource depletion is fishery over a period of five or ten years. clearly not serious, attempts to increase pro- Increased effort is accompanied by reduced duction can proceed without the immediate total catch or by very noticeable changes in necessity to conduct a more thorough the species composition of the catches assessment. Positive indications that the and/or the mean size of individual species resource is not in danger of overexploitation which are captured. If total catch has not include. 1) no overall downward trend in yet begun to decline, it will have remained landings for particular species or groups of more or less st.-ble despite increases in species with time, 2) a tendency for effort; thus, catch per unit effort (CPUE) increased effort (number of fishermen or will be declining in response to effort boats) to produce increased catches, 3) a ten- increases. Fishermen complain that they dency for increased effort to produce either can't catch as much as they used to; the fish the same or increasing catch per unit effort, are smaller and the higher quality species 4) no dramatic changes over time in the rela- are less abundant. In this situation, itwould tive abundance of different species caught be a mistake to encourage additional fish­ by a particular gear, and 5) no dramatic ing effort until a more thorough resource reduction in the average size of individual assessment has been completed. species captured by a particular gear over time. 3.2.4.2 Quantitative Resource Assessment A decision that a more detailed assessment was not urgent could be based simply on the More detailed resource assessments require observation that fishermen who have tradi- more specific quantitative information and tionally fished for particular species in a more sophisticated analytical procedures. particular area have no trouble catching all In this mode, resource assessment follows a the fish they want. It is important to bear in well established series of analytical steps, mind, however, that just because resources each of which requires estimates of certain seem to be abundant and there is no urgent parameters which are defined by a concep­ need for a more detailed resource assess- tual model. Two general types of models ment does not mean that a new assessment which are frequently used for stock assess­ will not be necessary sometime in the future ment purposes are: 1)the surp.us production as fishing pressure increases. Any decision model and 2) the dynamic pool model. Both to increase production from exploited or of these models are equilibrium mudels in underutilized resources should be accom- the sense that the size of the exploited stock panied by a commitment to collect the (which is a function of its rates of geowth, necessary quantitative data for a later more natural mortality and recruitment) and the complete resource assessment. yield which that stock will support are assumed to reach an equilibrium with any Other important elements of a preliminary given level of fishing effort or mortality. assessment could be 1) an estimate of the productivity of the ecosystem and the yield 3.2.4.2(a) Surplus Production Models that can be expected of important commer­ cial species according to their role in the Surplus production models require at least ecosystem and their life history strategy, four or five successive years of catch and and 2) the extent and location of accessible effort data for a given unit fishery (defined fishing grounds. Again, quantitative data in terms ofthe species captured and the gear are not required. It may suffice to know that used) which ideally represent a range from the current yield of a pelagic species which low to high effort. These models have been feeds on phytoplankton is only half (or dou- applied on a "total biomass" basis (all ble) what is to be expected on the basis of exploited species combined) and to individ­

20 in ual species populations. Surplus yJeld is the model more adaptable to situations considered conceptually as a function of which separate estimates of natural mortal­ population size (see Figure 2) and analyti- ity (M) and growth (K) are not feasible. cally as a function of effort. Model parame- Furthermore, the compilation of size infor­ ters are estimated from plots of CPUE mation and M and K estimates for individ­ versus effort. :' A common resource manage- ual species and unit stocks (Pauly, 1978; ment objective is to maintain MSY - a 1979) has led to the derivation of empirical defined by the peak of the yield curve equations which relate growth rates to max­ point to shown in Figure 2. Management is aimed at imum size and natural mortality rates adjusting effort in order to maintain a stock growth, size and mean environmental species or size which produces MSY. Surplus produc- temperatures for individual models are particularly amenable to stocks (Pauly, 1979; 1980a and b). Earlier tion demon­ either biological or economic analysis since work by Beverton and Holt (1959) rather than total catch may be strated that individual taxonomic groups total revenue by a examined as a function of effort and (families, genera) were characterized costs can easily be factored into the more or less constant M/K ratio, thus mak­ because K for an model (see section 3.4). ing it possible to estimate M or individual species belonging to a given tax­ 3.2.4.2(b) Dynamic Pool Models onomic group when an estimate of only one parameter is available. These procedures types, represent "shortcuts" for estimating popu- TheDynamic original pool modelmodels (Bevertonare of various and Holt, lation parameters which can be considered when available data are limited. surplus yield as a function 1957) considers recruitment and mor­ of the rates c growth, tality for individual unit stocks. Using esti- 3.2.4.3 Selection of an Appropriate mates of these biological rates of change Yield Model plus other parameter estimates derived far from age or size composition information, Each of the yield models mentioned so model predicts yield as a function of the has certain advantages and disadvantages the consi­ age or size when fish are first captured and associated with it which should be model fishing mortality. Maximum yield results dered when selecting an appropriate the proper combination of age or size for a particular resource evaluation. Simpli­ from case at first capture and fishing mortality. Man- fying assumptions which apply in each agement is aimed at adjusting the sizes of should be examined carefully and results fish captured (by changing mesh size in a interpreted accordingly. The use of a partic­ or a gill net, for example) or the ular analytical approach also depends on trawl orga­ amount of fishing mortality. In practice, the biological characteristics of the in fishing mortality must be nism(s) which is (are)exploited and the kind changes which can related to changes in fishing effort since of data which are available or there is no direct way to regulate fishing easily be obtained. 4 is evaluated ar, the weight mortality. Yield of the surplus produc­ which can be harvested from each individ- A major advantage of the tion model is that it does not require detailed ual recruit when reliable estimates is the population biological data; a major disadvantage number of recruits Pntering yield esti­ Tlh5 version of the Bever- that in order to make reliable are not available. data must be availa­ Holt dynamic pool model is called mates, catch and effort ton and years and include a range the yield per recruit model. ble for several from low to high effort. Also,the importance in the original Bever- of defining the fishing power of different Besides; modifications for Holt model which peimit the use of fishing gears and standardizing ton and time instead of age-speciic parameters changes in capture efficiency over size per (Holt, 1962), other modifications make it should not be underestimated. The yield to use estimates of mortality to recruit (Y/R) model requires a number of possible be growth ratios (Kutty, 1970), thus making parameter estimates, but analysis can

I There are two forms of surplus production model which consider CPUE as a linear or an exponential function of effort. by Schaefer (1954) and the exponential model by Garrod (1969) and Fox (1970). In The linear model was developed permits the fitting addition, Pella and Tomlinson (1969) proposed a general version of the surplus production model which of multiple functions to CPUE versus effort data.

4Fishing mortality is defined in terms of the probability that any individual fish will die as a resultof exploitation during a particular time period.

21 based on estimates collected during a single There are two types of exploratory fishing year. Two major disadvantages of the yield survey: 1) surveys which simulate commer­ per recruit model are 1) that direct regula- cial fishing operations and 2) surveys which tions of fishing effort are not possible, and estimate stock size or biomass. Simulated 2) recruitment is usually unknown. On the commercial fishing surveys are carried out other hand, if reliable estimates of all the on new fishing grounds (or using new har­ necessary biological parameters are availa- vest techniques) and are intended to demon­ ble, the Y/R model is a much more powerful strate whether or not commercial fishing on management tool since it allows for the these previously unexplored grounds (or separate regulation of size selective fishing using new techniques) is feasible. The objec­ gear (i.e. gear which captures fish in a cer- tive is to maximize catch rates by locating tain size range) and fishing mortality, the most productive fishing grounds (or by using the most efficient harvest technol­ 3.2.4.4 Resource Surveys ogy); no attempt is made to obtain a reliable estimate of the mean catch rate for the Exploratory fishing surveys conducted entire area inhabited by the stock. Catch aboard commercial fishing vessels or rates obtained from simulated commercial research vessels equipped with standard fishing surveys therefore provide an overes­ commercial fishing gear can produce esti- timate of total resource abundance. Surveys mates of resource abundance, distribution which are intended to produce reliable esti­ and seasonal availability. Other types of mates of stock size must be designed so that survey such as egg and larval surveys or individual catches are made at randomly hydroacoustic surveys may be applicable in selected stations over the entire survey area some situations for some species, but are not during all seasons of the year since resource as exploratory fishing abundance can be expected to change both as frequently applied ' surveys, particularly in the assessment of as a result of location and time of year. In it is critical that a standard gear valsmall-scale surveys fishery(see S, resources. ith and EggRichardson, and lar- typeaddition, is used and a standard fishing proce­ n a sa pling oce­ require a large number of samples to type is used 1977) is followed on each sampling occasion. overcome problems of extreme statistical dure variability and also require knowledge of Estimates of stock biomass can only be the fecundity (i.e. the average number of inferred from catch rates (weight/unit time) eggs produced per female of a given size) of if one knows: 1) the proportion of the total the species which are being studied, the stock biomass in a given area sampled dur­ identification of eggs and larvae (preferably ing a given period of time, and 2) the total by species), the exact timing of spawning area which the stock inhabits. Procedures and the rates of egg and larval mortality for estimating capture efficiency and for between the time of spawning and sam- converting catch per unit time data to catch pling. Hydroacoustic surveys (See Burc- per unit area data have been developed to zynski, 1979) are more applicable to some extent for trawls which sweep a schooling species whose identity can be known area of the bottom per unit time at a deduced from the nature of the acoustical given speed and which may or may not cap­ signal which is produced: verification is ture all the fish in their path. Trawl perfor­ usually necessary and requires the use of a mance depends on a number of factors such vessel equipped with some kind of appro- as the size and design of the net and doors, priate capture gear. The use of data mesh size, towing speed, etc. Since trawl gathered by remote sensing gear placed surveys can only be relied upon to assess aboard aircraft or even satellites can be use- demersal fish populations which inhabit ful for purposes of mapping sea surface bottom areas where trawls can be towed, temperatures or chhcrophyll concentra- they are not applicable for most small-scale tions, but direct applications for resource fishery assessment purposes. Wider appli­ assessment are not feasible at this point, cation of survey techniques to tropical Finally, it should be remembered that any small-scale fisheries will require the use of kind of resource survey, including explora- other gear types and the development of tory fishing, is expensIve. Further discus- suitable methodologies. Further discussion sion of resource surveys in this guide is of exploratory fishing methods will there­ limited to exploratory fishing surveys, fore not be included in this guide; the reader

IsIdeally, the survey area should include the entire area inhabited by thestock, or that proportion of thetotal area which is routinely fished by the commercial fleet. In practice, biomass can only be estimated for the area which was in fact surveyed, i.e. the area within which sampling stations were randomly selected.

22 is referred to Mackett (1973) or Saville (1977) have therefore been made to modify the con­ for more information. ventional models to include the effects of species interactions such as competition for Even in situations where trawls can be used food and space, predator/prey relations and to estimate biomass, estimates ofmaximum the exploitation of several - or many ­ sustainable yield are very approximate species by the same gear. At the same time, since they are based on extremely variable more attention is being paid to models estimates of mean biornass and single esti- which do not assume constant survival mates of natural or ti '; mortality rates rates throughout the life cycle and models which are frequently aijpAied to all species which allow for random variations in caught in the trawl, thus ignoring differen- recruitment, growth and mortality rates. ces which may exist between species. Although these represent important advan­ ces in assessment, the models are diverse, 3.2.4.5 Empirical Yield Models complicated, and require a great deal of even Another group of predictive yield models indata, developedmore thancountries, is generally notavailableto mention Even though this guide as empirical models. developing ones. may be classified data requirements and collection are models which simply establish describes These methods which relate to unit stock (single statistical correlations between variables, on meties asseltento it s ock (single They can be used to forecast catch based emphas­ data species) assessments, it should be 1) historical trends in catch and effort assessments. used for and an estimateand oofn efortineffortetimte in tethe u-coingizedup-coming multispeciesthat these data can also be between year and 2) historical relationship catch (or recruitment) and some set of envi- 3.3 SOCIOCULTURAL INFORMATION ronmental variables. These empirical mod­ els make no attempt to explain the 3.3.1 Introduction which produce changes in mechanisms II, sociocultural yield or to establish causality. Computer As mentioned in Chapter can be used to evaluate the sen- information performs two functions: it facil. simulations and eco­ sitivity of yield predictions to changes in itates the collection of biological variables and combinations of nomic data and is a key element in making individual decisions. variables. These variables can include eco- development and management sociocultural factors as well as In this chapter, the types of information nomic and into three biological ones. Because these models needed are identified and divided data and data processing general categories. Individual data needs in require so much ref­ they are usually not practical for each category are listed separately and capability, Chapter evaluation of resources exploited by erenced according to the section in the detail. small-scale fisheries in developing IV where they are described in countries. 3.3.2 General Background Information 3.2.4.6 Multispecies Fisheries General sociocultural information is the struc­ Considerable attention has been focused in required in order to understand years on multispecies fisheries and ture and function of a small-scale fishery recent If interven­ multispecies stock assessment. Conven- system (see Figures 3 and 4). such as the Beverton-Holt tions in the fishery are to succeed, this infor­ tional models basic and surplus production model were mation is crucial. Perhaps the most model is needed developed originally for application to sin- sociocultural information which populations. Since no species concerns the identification of groups gle species to belong to exists in isolation from other species in the (4.2.2.1). If fishermen appear and since many distinctive groups (linguistic, religious, or fishery or in the ecosystem, to gears exploit more than one - and ethnic groups, for example) it is necessary fishing located and the often many - species at once, attempts determine where they are

population from estimates of "Gulland (1971) proposed the following equation for estimating MSY for an unexploited biomass and natural mortality: MSY = 0.5 M Bo where M =instantaneous rate of natural mortality and B = the maximum (unexploited) population biomass as estimated from surveys. This equation has been modified for use with exploited populations, i.e. MSY = 0.5 Z Bo where Z = instantaneous rate of total mortality B = the exploited population biomass.

23 degrees of intergroup tension which exist 3.3.4 Innovation, Occupational Prefer­ between them (4.2.2.2). Often, the design of ence and Training development projects will be influenced by intergroup differences; if groups are identi- Information concerning previous attempts fled during early stages of planning, it will to introduce innovations, their type, and be possible to more realistically estimate their acceptance should also be gathered at project costs (Cochrane, 1979). this time (4.2.4.1) as a means of evaluating the potential for change as well as the most After groups have been identified, the effective means for accomplishing it. Atti­ numbers and locations of potential project tudes toward risk, change, and investment participants within these groups must be (4.2.4.2) can also be evaluated at this stage determined (4.2.2.3). At this time it is also of the project. These will be important varia­ possible to obtaip information concerning bles to be considered in the cost/benefit traditional communication channels analysis of the various technological (4.2.2.4) which will facilitate other data col- alternatives. lection programs as well as enhance com­ munication of proposed changes (Pollnac After the most appropriate technology is and Sutinen, 1980). Information needs at selected, it is important to assess how the this point include 1) level of develorment of expected changes which will follow are per­ target communities (4.2.2.5), 2) occupa- ceived by individuals and groups which will tional structure of the target region (4.2.2.6), be affected. Often the new technology is 3) numbers of small-scale fishermen, both only one of many changes which are consi­ employed and unemployed (4.2.2.7), 4) avail- dered. For example, if the costs of the tech­ ability of alternative occupations for fisher- nology require that credit be provided, then men (4.2.2.8), 5) temporal distribution of the expected methods for extending credit fishing effort, (4.2.2.9), and 6) local knowl- also need to be evaluated prior to introduc­ edge about fishing and fish (4.2.2.10). tion (Pollnac, 1981). Ifnew fishermen are recruited to the uccupa­ 3.3.3 Social Structure of the tion, attitudes ofnon-fishermen toward fish­ Occupation ing should be assessed (4.2.5.1). If potential new fishermen require training and if there is a pool of individuals who wish to become If a decision is made to increase production fishermen, then social and economic costs by improving technological efficiency and benefits of alternative training tecb­ rather than increasing the number of fisher- niques should be compared. This anaiy.ais men, then information needs indicated by should recognize the importance of iusing Figure 4 become essential. Basically, what locally acceptable training techniques for is required is an analysis of the social struc- adults. It has been reported that training ture of the occupation of fishing and its efforts have failed for reasons as trivial as place in the social structure of the commun- inappropriate setting (Foster, 1973). Both ity. The required data include 1) fishing the recruitment and training of new fisher­ gear types and ownership patterns (4.2.3.1), men should be treated as innovations 2) crew size and social composition (4.2.3.2), requiring information of the types listed 3) criteria for crew selection (4.2.3.3), 4) below. degree of occupational mobility (4.2.3.4), 5) types of interaction between crew members In the situation where resources need to be and between owner and crew (4.2.3.5), 6) managed, information referred to above degree of on-and-off vessel occupational spe- (such as alternative occupations for fisher­ cialization (4.2.3.6), 7) relationship of fish- men and skill levels) will make the selection ing groups with other social groups in the of appropriate management procedures eas­ society, (4.2.3.7), and 8) the local distribu- ier. It will, however, be necessary to obtain tion of wealth 7 and power (4.2.3.8). This information concerning fishermen's atti­ information is also valuable if the number tudes toward fishing as contrasted with pos­ of fishermen in the community increases, sible alternative occupations (4.2.5.2) as but technology remains relatively well as attitudes towards changes in income unchanged. (4.2.5.3). Proposed management strategies

7Wealth refers to different things in different societies (Cochrane, 1979; Foster, 1973). In some, a wealthy man is one with many r'lildren, in another, wealth is equated with the leisure time to fulfill social obligations, while in others it is equated with the possession of certain material things (e.g., gold, religious statues, yams, cattle, etc.).

24 (such as gear restrictions, closed seasons, or In this section of Chapter III we shall look size limits) must also be treated as innova- more closely at economic forces at work and tions requiring the specific information see why information about them is impor­ listed below. tant. The qualitative results of using simple important bioeconomic relations will point out infor­ As changes are introduced, it is to which the following mation needs. We shall see that attempts to evaluate the ways in to conserve of the changes are perceived by help fishermen and consumers, attributes and to rationalize the harvest fishermen: 1) complexity (4.2.5.4); 2) the resource the are often at odds with each other. compatibility (4.2.5.5); 3) relative advan- sector tage (4.2.5.6); 4) trialability (4.2.5.7); and 5) While the economic operation of the entire observability (4.2.5.8), (Rogers and Shoe- system connects suppliers of inputs to fish­ maker, 1971; Pollnac, 1976). Individual ing at one extreme and the consumers of attributes associated with innovativeness fish at the other, the primary focus of what (4.2.5.9) should be assessed at this time as a follows will be on the economic cperation of means of identifying individuals who will the harvest sector itself. For the purposes of serve as appropriate subjects for demonstra- this guide we define this to be limited to the tion purposes. purchase ofinputs by fishermen, the fishing or harvesting process and the sale of output to middlemen or consumers (see 4.3.1). It is 3.4 4ECONOMIC INFORMATION not our purpose to explain the analyses commonly used to obtain this 3.4.1 Introduction which are Detailed explanations exist in the more important reasons for information. Some of several sources and they will be referencad. acquiring economic information about the fishery system can be stated in general 3.4.2 The Bioeconomic State Of terms. They include the need: Harvesting 1) to understand how the system functions; 3.4.2.1 Open Access the potential of the fishery An open access fishery is one in which there 2) to evaluate for to provide protein and employment; is freedom to enter and exit - freedom leave it and to evalu- an individual to take up fishing or 3) to be better able to suggest available to impact ofinterventions, i.e. depending on the opportunitie- ate the potential of the opportunity this free­ of regulations and/or investment projects; him. Because dom offers, there is a tendency for open 4) to determine the impact on the fishery access fisheries to overexploit the resource of development in other areas- especially in in an economic sense. There is a tendency the development of the industrial fishery. from society's, and not the individual fisher­ In Chapter I we examined a series of deci- man's, point of view for more labor and capi­ sions that would have to be faced in tal to be used in catching the fish than is attempting to expand the fishery's role in desirable. New investment persists until the supplying protein. A significant amount of total cost of catching the fish is equal to the information would have been required to total revenue the fishermen receive for all determine if that goal were realizable, if the the fish during that year. This may occur to interventions proposed would have the the left of, at, or beyond MSY (Figure 2). We desired result, and ifthe entire process were will examine how an open access fishery worthwhile. Any particular intervention develops around a newly exploited resource would require information from many sour- and how individual decisions lead to eco­ ces and the coordination ofmany ministries nomic and perhaps biological overexploita­ or departments. We concern ourselves here tion. with intormation (this chapter), data (Chap­ ter IV) and collection procedures (Chapters When a fishery is new, fishing represents an V and VI) particular to the fishery depart- opportunity for people to make better ment's ongoing responsibility for small- incomes than they would normally. They fisheries. This information forms a enter the fishery in anticipation of receiving scale their core for project design and evaluation. For better returns (or at least as good) for information on fishery investment labor and invested money than they would detailed or projects see Campleman (1976) and Eng- receive in the next best employment strom (1974). investment opportunity. RIf local, regional educational opportunity here to mean those which can be taken advantage of. There may exIst barriers, cultural, "We use are no such barriers to or social - which prevent opportunities from being pursued. We will assume for now that there entering the fishery.

25 or national economic conditions are such the dotted lines in Figure 6. Note that the that current incomes are low, then the "nor- equilibrium level of total revenue (and there­ mal" income may mean subsistence farm- fore catch) reached at point Ai is above that ing, rural or urban unemployment. If we reached at C. Nevertheless, at point At the assume that the selling price of our single annual total reventie of the fishery is equal species of fish remains constant regardless to the annual total cost. of how much is i old to middlemen and that Figure 7 compares the annual revenue and all fishing ope'ations are the same, then costs of an average lone fisherman at points Figure 6 represents a simplified picture of C'and Ci respectively. The operating costs two of the many possible alternate courses are the same under both situations; how­ of a new open access fishery through time. It ever, the incomes differ substantially. At C' shows the annual total revenue (catch mul- the average fisherman's income isabove tiplied by its price) received for fish and the "normal." Itiscomposed of two parts: the annual total cost of catching the fish (effort return which he would be able to earn in multiplied by the cost per unit effort) at dif- other opportunities (his opportunity cost of ferent points in time.9 We will first discuss fishing) plus a share of the industry-wide situation C, represented by the solid lines, profits being earned (at C' total revenue Initially there is a large quantity of fish to exceeds total cost). At equilibrium point C, be caught by relatively few fishermen. This however, these profits are gone. The aver­ quantity is even larger than the quantity of age fisherman, given our assumption of surplus production at MSY because fisher- similar firms, is earning exactly the oppor­ men are reducing the initial total popula- tinity cost of his time and the opportunity tion as well as catching some of the surplus cost of his equity capital."' production. As more people realize the The existence of the fishery-wide profits at opportunity for larger incomes, they enter C' as at all the points before C', drew people the fishery. As more and more begin fish- into the fishery. The increased effort on the ing, the catch per unit effort begins to resource gradually diminished catch rates decline. There are still above "normal" and, consequently, the revenues earii fisher­ at, say, C' but less so returns to be made People continued to enter the than there were earlier. At C", for the first man received. revenue of the entire fishery doing so. time, the total the is just equal to the total cost of catching fish. At C" the average fisherman is making Points like At and C can be thought of as only a "normal" income. Because of the points on the equilibrium yield (total delay in the apparent response of the stock revenue) curve. Figure 8 shows two bioeco­ to -fishing and because of the difficulty in nomic equilibrium points and their asso­ selling boats and gear, there are years of ciated total cost (of effort) lines (cf. disequilibrium in which the total costs Anderson, 1977). Where an equilibrium exceed what the fishermen receive for the point wiil lie relative to MSY depends on the fish (e.g., point C"'). After a while, in our reproductive nature of the resource, the fish­ simplified fishery, a somewhat stable equili- ing mortality inflicted on the resource by a brium is reached (at C). nominal unit of effort, the price received for The initial rate of response of the entrants theFigure fish 7 thatand thethese cost costs of includeeffort. Recalla "normal" from depends upon the perceived above normal Fgr htteecssicuea"oml resource has alow returns. If their incomes had been above return. Ingeneral, ifthe at o fh in alityor he methodsubsistence of fishinglevel, forusedexample,were moreand costlyif the a largerge impact on fishing mortality, or the mthod thecajusoin g bede , e orescosy price of fish is high and the cost of effort than in thae e just described, the response (including the normal return) is low, then might have been quite different. An exam- the bioeconomic equilibrium point is more ple is given by situation A, represented by likely to lie beyond MSY.

9We use "effort" here to mean nominal effort such as total nunber of days fished so that there is a correlation between nominal effort and the number of people in the fishery. The cost per unit effort is assumed constant but differs between A and C.

"' Economists refer to this "normal" income available in other occupations as the opportunity cost. The opportunity cost of fishing is the foregone returns available in the next best use of a person's labor and money. Opportunity cost is a useful measure which helps explain why individuals switch occupations and/or investments. As such it can be used to predict, for example, the supply of labor to an indus try. The absence ofalternate employment and/or investment opportunities is an indication that people will persist in their woi #'inves~m ,nts perhaps in spite ofpoverty level incomes. Smith (1979) examines this in relation to small-scale fisheriei. research issues.

26 Annual Total Revenue TRc­ or Catch ! and Annual M- Total Cost TCA C1

TIME

Figure 6. Two possible courses ofexploitation of an unregulated fishery. In situationC thepriceof fish is higher,the cost perunit of effort is lower andentry ceases laterthan in situationA. The equilbriumcatchin situationA, A , is closerto MSY thanthe equilibriumcatchin C, Ct. Neither equilibriumis perfectly stable.At both Ai and C industry-wideprofits arezero.

< RETURN TO LABOR > jI

- < RETURN T O CAP ITAt > T T 0 1? 0 T T TT AO oA

L T VAR IABLE COSTS TL A A R L LR E E EO OE NS SN UT TU E F IXED COSTS E

Figure 7. Annual total rvenue aid annual total costs for an average lone fisherman in situation C. At C' the fishermanearnsgreaterthan "normal"returnsto his invested money andlabor.At C his revenue is equal to his costs. Hiscosts include hisperceptionof the opportunitycostof his money and time. All costs are considered in this example to be independentof the value of the catch. Otherwise the costs at C' would be greaterthan those at C (see 4.3).

28 argue Let us assume that the fishery is in a state of in the number of fishermen. We might and economic overexploitation that some fishermen find that this biological below (point Ci in Figure 8). Ci is undesirable in a increased cost reduces their income sense because surplus production their opportunity cost of fishing, i.e. they biological offer a less than it could be. The equilibrium have alternate opportunities which is the fishery. stock size is likewise smaller and therefore larger return; hence, they leave to the effects of environ- The actual response is likely to be much more susceptible a group, and/or recruitment variations. This more complicated. They might, as mental or substitute at Ai. The economic conditions at fish closer to their home ports is not so It is safe to say, Ci are undesirable, but this holds for those sail power for motor power. In society's view too many however, that i.nasmuch as fuel use is corre­ at Ai as well. fuel use resources are being used in both lated with effective effort, less productive on the resource. cases to produce fish. would mean less pressure The long run result of either explanation Let us examine the reaction of a biologically (shown as C )is that the stock size increases and economically overexploited fishery to a and the flow of fish to market increases. As change in 1) the price of fish; 2) the price of we might expect, the remaining fishermen an input (fuel) and 3) a technological are no better off than before. The mix of change." If, after a campaign promoting capital and labor may have changed but, as fish consumption, the price of fish were to shown in the figure, the to'al amount increase, the total revenue curve in Figure 9 invested will have increased. would shift up (new pice multiplied by appear between An increase in the cost of catching a given catch). A vertical gap would about cost at the old equili- quantity of fish can also be brought the new curve and the regulations Fishery-wide profits would by the imposition of biological brium, Ci. size at first capture. available again. To take advantage such as those that limit become cause the temporary adop­ of this profit, fishermen would fish more Such regulations fishery. tion of harvesting inefficiencies. They and new fishermen would enter the The would fall and achieve some biological good however. Catch per unit nominal effort adoption of would gradually disappear. long run effect of mandatory those profits mesh size nets can be seen a period of adjustment a new equili- larger minimum After 11 and 11a. The short run effects brium will be reached at C2. At C2 the in Figures in both are more clouded in this example than in resource is even more overexploited Because men and money in most of these long run scenarios. senses. The amount of reduced, people larger than before and the fish the catch per day is initially the fishery is fishery immediately or smaller. Figure 9a shows the may either leave the population fish longer hours. The increased mesh reduced flow of fish as a result of the price try to indication that size being used by fewer total fishermen increase. An approximate less pressure on the been reduced may be obtained results in significantly the catch has both the numbers and by dropping a vertical line from the new resource. Gradually old sustainable size of the fish caught increase. Ultimately equilibrium C2 down to the the average fisherman profits reappear and people enter total revenue curve. The level than at Ci. In the long fishery. Profits exist even at the former at C2 is no better off because the paying more for less fish, of nominal effort simply run consumers, had an opportunity to rejuven­ are worse off. resource has ate itself. The level of nominal effort and ia show the long run increases until the profits disappear at the Figures 10 compli­ response of the entire fishery to an increase new equilibrium C. In a sense this to an in an input cost. If we assume that the input cated set of dynamics can be compared is fuel (though it could as well be ice or bait), imposed form of savings: current consump­ then the average cost of a day's fishing tion is delayed so that a larger (sustainable) would increase. We indicate this in our pic- supply a available later. At C4, however, in the ture of the harvest sector by rotating the there is a larger investment by society total cost curve up and to the left. A tempor- fishery and, perhaps, more people fishing. ary condition is immediately created in And while the fljw of fish to market has which the total cost of fishing exceeds the stabilized at a higher level, the individual total revenues. Given that we are using a fishermen, as in all of the situations des­ nominal measure of effort, the figure shows cribed above, have not benefited. Industry­ a reduction in the number of days fished or wide profit is zero. at hand, will recognize the limitations imposed by the straight line total cost function. For the purposes I Economists text will discuss differential responses to the changes in the examples will be shown as straight line changes while the impacts, suggesting different opportunity costs and inframarginal rents.

29 Equilibrium Total Revenue or Catch and / Equilibrium /EMSY or Maximum Total Revenuec Total Cost of Effort $, Tons at C, and Ai

EFFORT EA, Eus Ec,

Figure 8. Two cases of economic overexploitation. At Ai there is no biologicaloverexploitation while at C there is. In both cases industry-wideprofits arezero. The per-unitcost of nominal fishingeffort is much greateralonglineA thanit is alongline C. The flow offish to consumersis the same in each case. Equilibrium Total Revenue .. . and Equilibrium Total Cost C2 of Effort

I

C,05*0

Eci Ec2 EFFORT Catch in Tons Ci

TIME

Figures 9 and 9a. Long run response of an economically and biologically overexploited fishery to directlyabove an increase in the price offish. In theshortrun,industry-wideprofitswould appear C attractingnew fishermenandincreasingthe amountoffishingby thosealreadyin thefishery. will eventually be establishedat C2 with greateroverfishing andgreater A new equilibrium be overinvestmentofpeople andmoney in thefishery.In addition,the flow offish to market will less at C2 (Figure9a). Equilibrium Total S.Tons Revenue c3 and Equilibrium Total Cost of Effort S. Tons-

EFFORT Ec3 Ec,

Catch in Tons C3

C,

TIME

Figures 10 and lOa. Long run response of an economically and biologically overexploited fishery to an increase in the price offuel. In the shortrunthere will be lesseffort expendedon thestocks as fishermen either leave the fishery, fish less, fish less far from home, or switch methods of propulsion.The stock will recoverto some extent andcatchperunit nominaleffort will increase. A new equilibriumwill be establishedat CO with zero industry-wideprofits but a greaterflow of fish to market (FigurelOa). Equilibrium Total Revenue and Equilibrium Total Cost of Effort C S; Tons C4 $, Tons C,

EFFORT Eci Ec4 Catch in Tons C4

C,

TIME

Figures 11 and Ila. Long run response of an economically andbiologically overexploited fishery decreaseas to a larger minimum mesh regulation. In the shortrun thepressureon the stocks will efficient gearcatchesfewer andon averageolderfish.As thestock regenerates,industry­ the less the fishery. wide profits will appear.New fishermen will enter orformer fishermenwill reenter The new equilibriumat C4 shows a greateramountof nominaleffort employed, largersustained yield (Figure11a) and largerbut equal total revenue and total cost of fishing. Without redrawing Figures 9, 10 and 11, we The owner would incur management costs, can see the results of the reverse situations ­ of course. There would be research costs, falling prices, decreased costs of nominal administrative costs, negotiation and legal effort, and more biologically effective gear ­ fees, and substantial enforcement costs. by starting with the new equilibria and These would differ depending on which working backward. We should keep in mind method was chosen to collect the rent. After that all of these equilibria are of a long run a few, initial, unprofitable years, he might nature and are the theoretical end result of be able to finance these costs from the rents the shorter run dynamics of firm and stock received. These rents would increase as the adjustments. stock replenished itself from the open access point Ci through MSY to MEY. The process 3.4.2.2 Resource Rents and Owners of reducing effort might take several years The "industry-wide profit aoeporu-involved,above opportun- and, dependingthe managementon all the various costs factorsmight "resource rent." exceed the rent at MSY and, possibly, even ity cost" is often called the owner of the most easily thought of as the at MEY. If there were an This rent is that an owner of theatMYIfheewrano erfte total annual amount resource, then, and if he decided to collect resourceteoa ane coulduld chargeham ththose noner using og histhes the rent due him, he would have to estimate rent and compare it to In some traditional societies in the the magnitude of the stocks. management costs. Pacific, fish resources are "owned" by cer- the estimated tain coastal communities. These communi­ ties lease rights to fish to others in exchange Whether there exists such an owner is a for goods and services, legal, even constitutional, question which For the purposes of explanation, let us has implications for the disposition of the imagine a resource owner and assume that rent and for the magnitude and distribution he confronts a situation in which there is of management and certain other costs. If both biological and economic overexploita- society is considered the owner and the tion ofhis resource. If, starting at a position government can and does act on its behalf such as C,in Figure 8, he were to charge a in collecting the rent, then broader effi­ small fee per pound of fish caught, some ciency and equity questions are raised. The fishermen would be forced to leave the narrow efficiency goal, from society's point fishery; their costs would have risen while of view, is to produce ihe biologically and the price they received would have economically proper amount of fish at min­ remained the same. After a while the owner imumcost (Anderson, 1977). This minimum would have a somewhat larger stock of 'ish cost refers to the opportunity cost of scarce than existed at Ci and fewer fishermen seek- resources used to harvest the fish. Trying to ing to fish. If he were to continue each year bring about the proper level of harvest has to raise the price per pound caught, effort costs of its own. We have already discusseu would be reduced further. Consequently his management costs. But perhaps additional resource would replenish itself to a level costs will be incurred by society, such as where it would be harvestable at MSY. At payments to fishermen and their families MSY he might be able to collect a considera- from established social insurance pro­ ble annual rent.12 There would also be sig- grams. Or, perhaps, some other costs will be nificantly fewer fishermen fishing. The assumed by society out of equity considera­ owner could pursue this policy of increasing tions. Perhaps some of the rent should be the catch fee until he maximized his rent at used to buy back (retire) boats and gear MEY. gradually. Or perhaps alternate income for those would come about if the opportunities should be developed The same results for effort, resource owner taxed or charged as long as the effort was, in turn, directly From quite another viewpoint, arguments related to fishing mortality (i.e. effective can be made for recognizing existing fisher­ effort). Likewise, he could hold an auction men as owners of rights to a certain percen­ and have fishermen bid for the privilege of tage of the annual catch. This total amount fishing for and catching a certain amount of could -Iadjusted annually. For example, fish. In other words, he would have the enforcement costs could be reduced by option of controlling the amount taken c -ach establishing individ'al rights which would year either directly or indirectly (effecive be recognized in the courts. The number of effort).

12The amount depends on the factors we are familiar with: the price of the fish, the productivity of the resource and the costs of fishing.

34 is fore­ fishermen being granted or being recog- (but which is not costless to collect) as having these rights could vary gone for all uses; and, as we will see below, nized than it could from the number currently fishing to some the price of fish will be higher smaller set based on, say, historical partici- be. pation. Once granted, these rights might be freely sold, traded, or gradually bought 3.4.2.3 Other Options back by the government. Having seen the case of biological and eco­ examine an There is considera us middle giound where nomic overexploitation, let us the resource society and fishermen can share rents. It is open access situation in which that many, if any, management is not as yet fished beyond MSY. We will doubtful in equilibrium at schemes would be able to succeed without assume that the fishery is to the fishermen. We would some point like point Ai in Figure 12. The incentives Ai is larger expect the enthusiasm of all fishermen to be standing stock which can yield any scheme if they were than that which yields Ci, larger, in fact, diminished for MSY. The asked to trade the certainty of increased than that which could produce is the same in either costs today for uncertain future returns, flow of fish to market case but less than the flow at MSY. Given The issuea involved in trying to improve the our assumptions, there are fewer fishermen condition of an economically overexploited at Ai than at C but there is no industry-wide open access fishery are many and compli- profit to share in either case; the incomes of cated. Several different management those fishing are the same. schemes are in ise and many more are 3 In general these Is there economic overinvestment of men under discussion.' well? Yes. recognize society's interest in see- and money in this situation as schemes at Ai as it the resource is properly used (nar- Total revenue equals total cost ing that way Figure 12 is goal) regardless of the ownership does at Ci. In fact, given the row and total costs are question. They then examine the broader drawn, the total revenues aspects of who, among approximately the same (we measure these efficiency and equity axis, not consumers and fishermen, benefits values by distances on the vertical society, amount of overin­ and who pays the costs. by areas). However, the vestment in situation A, OIA, given the tech­ We saw earlier that the imposition of biolog- nology and its per unit cost of effort, is less ical restrictions which do not properly con- than that in situation C, OIc. This may not trol total effort can only address some be immediately clear, especially since the aspects of the problem. While the quantity individual profitability of the fishing units of fish flowing to the markets will eventu- is the same in both situations. However, we ally increase as the stock rebounds, man- can see that the technological change that agement costs are incurred, excessive reduced the costs of fishing (per unit of nom­ resources (the money value of vessels, gear inal effort) has benefited the nation and the and labor) are still diverted from more pro- fishermen only in the short run. In the long ductive areas of the economy, and those run, there is a greater amount of rent fore­ who have not been forced out of the ;ishery gone at Ci than there was at At. off than before. There is, of are not better example to the option of doing nothing. This Depending on our goals, for course, the resource or to bet­ option is not costless, however: the eco- increase utilization of of the fishermen will per- ter the conditions of fishermen, the creation nomic condition a tool or as an sist and very possibly grow worse; the of rent can be viewed as either which may already be stressed, end in itself"I. The rent is, as we have seen, resource, revenue to will be increasingly exploiLed; the benefits the difference between the total protein supply from the the industry (to all fishing units) less total of an increased in the short run fishery are foregone; an annual amount of costs. Rent can be generated resource rent or industry- by increasing total revenues. Programs to money called the an export profit which is equal to the annual stimulate demand or to establish wide price received by value of the overinvestment in the fishery market can raise the

Some of the contribu­ international conferences have focused on the merits of different management schemes. 1."Several (1979) for example). tions to theee conferences are referenced at the end of this guide (see Pearse

for evaluating the the amount of rent actually or potentially available is a useful quantitative measure 14 In a fishery in the fishery, iicome levels weights to be given to non-efficiency considerations such as levels of employment alternate potential rent. and income distribution. MEY is simply the yield/effort combination of maximum

35 Equilibrium Total Revenue or Catch and Equilibrium Total Cost of Effort S. Tons MEY C

S, Tons MEY A I

EY A c EI"i EFFORT

Figure 12.Two casa of overinvestment. Similarto Figure8. CaseA shows no biologicaloverex­ ploitationwhich caseC does.An investment in more efficient technology reducestheperunitcost of nominal or effective effort from A to C. In the short run the returns to thatinvestment (the benefits)will bepositivebut ultimately,at C industry-wideprofitswill be zero. In additionthere is,at C, biologicaloverfishing,andthe loss ofagreateramount of potentialrent. Viewed another way there is a greateramount of socialoverinvestment in Case C (OIA versus Olc). total costs will have be weighed against 1) the amount of over­ fishermen. Reducing grown the same rent generating effect. Rent can investment in the fishery will have available by allowing access from RA to Rn and 2) the economic condition also be made at to new, perhaps more distant stocks of the greater number of people fishing through new vessel designs, for example. Bi. all of our equili­ This, in effect, redefines by this potential rent Control of the quantity of fish captured, brium yield figures. Once effective and the utilization of the whichever of the economically is available, effort stops the dissipa­ will begin to increase as shown in methods discussed above, resource of the cost lines at a 6. tion of rents along any Figure point short of intersection with the total are many situations in which indi- revenue curve. For any of the given situa­ There the respec­ genous barriers may slow the desired tions this rent is largest where biological exploitation. On the tive solid total cost lines in figure 13 end; it increased as it other hand these barriers can work in our diminishes along the dotted lines favor by substantially delaying the loss of approaches the total revenue curve/Without and height of the newly created rents. Many development real information on the shape aimed at increasing incomes to the total revenue curve, the cost curve, the projects up the fishermen have been premised on the exis- mix of capital and labor which make sharing rents, tence of cultural, educational (skills) or ge- effort and on the method of access) barriers which etc., it is impossible to reach anything more ographic (isolation, of will permit improved incomes for those than a general, qualitative evaluation for long periods of time. In the long any action. For example, an investmciit involved Point Ai to run, however, without some measure of con- that moves us from open access industry-wide profits will disap- point B2 results in less fish to cwisumers, trol these and the pear. The speed at which this will occur probably fewer fishermen fishing, rent (less depends on the nature of the barriers. creation of rent RB. If all of this management and other costs) is retained by Figure 13 shows various long run, equili- the government, the remaining fishermen's brium states of the harvest sector. The three incomes will not change. line.s r:epresent three different sets straight Gear of total costs of nominal effort -each repres- 3.4.2.4 Many Species and Mixed enting a different cost per unit of nominal cost is We have used several simplifying assump­ effort. Along line OA this per-unit Two it is lowest. We have seen tions in this static, long run analysis. highest, along OC species of out­ the siturtion at Ci. At Ai and of these implied that a single and discussed was captured by applying Bi we encounter economic but not biological put called "fish" and the effort produced by similar fishing units overexploitation. At MEYA, MEYB, of economically or firms. Tropical small-scale fisheries, MEYc we have three different, of species Given our narrow effi- course, exploit a large n umber optimal situations. combinations of boat, goal - the biologically and economi- using a variety of ciency and men. cally proper amount of fish extracted with gear the smallest total extraction cost - the These different fishing units or firm types situation at MEY is better than that at differ from each other in their effect on the MEYA and is "best" at MEYc. 15 resources as well as in their economic opera­ a different but often goal of investing in more effective tion. Each type catches The subset of all the species availa­ gears, for example, can be characterized as overlapping to some point .ble and has varying degrees ofcontrol over trying to move from line OA, species. point on OC. Without directing the effort at particular on line OB, even to a on a particu­ these investments will, The effective effort each exerts some controls as does the average size of or rapidly, lead to a new intersection lar species differs slowly catches. with the total revenue curve at, for example, fish each Bi or Ci. While a change from Ai to Bi gets In fisheries which exploit a very large more fish to the 111arkets - assuming the number of species, prices are often assigned rest of the delivery system is operating well - the benefits of increased production must to groups of species or togroups of species of

for each different a different MEY, and, therefore, a different economically proper quantity of catch 1.5There is obviously to the curve is equal to the slope of effort cost. The MEY point is graphically defined where the slope of a tangent per unit in the fishery is equal to themarginal (or the total cost line. It is here that the marginal costs of, say, an additional boat additional) revenue of that boat to the fishery.

37 Equilibrium Total Revenue or Catch and Equilibrium Total Cost of Effort MMEC $ Tons MEY C

$, Tons MEY 9 MEY 8R. $,TosMEY A Re.,/ 00. MEY A RA.. Tn "I.,"B,"M[ B 2 151

EMEY A EmFX B E.EY C EFR EMl E81FFR Eci Figure 13. Three different cost-per-unit-effort cases. Each has its own point of maximum eco­ nomic yield (MEY) where the potentialrent is largest.Beyond each of these points (dottedlines) the potentialrent (RA, RB, Re) diminishes untilitis zero as the cost lineandrevenueline intersect at A i, Bt, or Ci. An investment project or the independentadoptionof more efficieni technology which brings the fisheryfrom situationA to, say, B2 on line B will maximize socialreturns. The revenue will be at the level MEYa and the cost at the vertical level of B2. a certain size. These groups may or may not is very likely, management by fishing unit have similar biological characteristics. The types. The point here is that information commercial classes are generally assigned which is important to, and important relative prices on the basis of consumer because of different concentrations of peo­ taste preferences. These relative prices are pie and their fishing activity, should not be generally stable but may vary during cer- obscured. tain seasons as the quantity caught and species mix change or prior to traditional 3.4.3 The Delivery System events such as Easter. During these periods the price of all fish can change substan- Another assumption used above, that of a tially. Consequently there is a difference in constant demand price for fish regardless of revenues received by each firm type and by the quantity supplied, rarely holds. There all firm types over the course of a given year. are many more opportunities for the price to vary as both the quantity supplied and the Each type of fishing unit uses different quantity demanded change. Let's examine amounts of capital and labor and in differ- the flow of fish through the system and see ing combinations. This results in different what factors determine the price. sets of costs and, therefore, in different responses to changes in input prices. Some Figure 14 represents an abstracted fish pro­ variable costs are closely related to the duction and delivery system. Three markets amount of nominal effort exerted; others, are involved. In the first market variable particularly labor costs, can be more closely inputs (ice, bait, labor) and fixed inputs tied to the amount of fishing mortality (motors, gear)are supplied to the fishermen. inflicted (effective effort). Tie fixed costs of These are offered to the fishermen by suppli­ the different fishing units reflect, in large ers who are incurring costs to do so. The measure, the amount of capital invested, suppliers pay rent on buildings, import They include interest payments, deprecia- duties, and the opportunity cost of their time tion and some proportion of maintenance and labor. The fishermen purchase these and repair costs. inputs, combine them with their own labor, risk and opportunity costs, and convert As a result of all of this variation, the com- them into fishing effort. They anticipate posite measure of this biological and eco- that this process will result in output (fish). nomic activity, the firm's profit, differs over Their demand for inputs like ice and bait is firm types.'6 The profitability within a firm derived from the anticipated sale of fish to type can vary substantially over individual middlemen. firms, reflecting different skills or experience. In the second market the middlemen pur­ chase, transform and transport fresh fish A particular boat type, kind of gear and and incur costs from using ice, trucks, build­ crew configuration is often used by all ofthe ings, freezers, from assuming risks, and fishermen of a particular community, sec- from foregoing other opportunities for their tion of coast or island. Likewise, the set of labor and money. The demand of the mid­ alternate opportunities available to fisher- dlemen for fresh fish from the small-scale men is, to a greater or lesser degree, condi- fishery is derived from an anticipation of tioned by the economic circumstances of revenue in the third market, i.e. from the their community. Therefore changes in the sale of their processed product to fishery, whether induced or not, have an consumers. impact that is neither geographically nor socioeconomically random. Information If we assume away any spoilage, we see that about different biological, social and eco- the quantity of fish sold is the same in nomic characteristics of the fishing units' markets two and three. The prices asked by operation and their setting should be avail- the middleman are higher per unit so that able prior to any attempt to intervene, he can cover his costs. of both the total Furthermore, management the biomass and its value in a multispecies The quantity of fish that flows from fishery will involve management of and, it harvest sector to consumers in a period of

and profit can be thought of as the annual accounting profit, i.e. as the difference between revenues and fixed ' The firm's cost of labor (the variable costs. If we were to add to the fixed costs of the firm an amount equal to the firm's opportunity or subtract this amount from the firm's profits, then the differences across firm types would be much owner's) and capital, the costs for If, after having added each firm's opportunity cosls to its fixed costs, the revenues still exceed all of smaller. new entrants. many firms, this difference is an indication of the existence of industry-wide profit which attracts

39 P CO rE F FT F 1

__Z LT Si

MARKET I" INPUTS

P

PI MARKET 2: FRESH FISH

al CL(KG. )

P

P2

MARKET 3: FINAL PRODUCT

O((KG.)

Figure 14. A simplified annual three market delivery system. "hefishermenpurchaseinputsand produce nominal ,[fort. This results in quantities of catch depending upon the size of the availablestocks offish. Notice thatthe quantityoffishpurchasedby primarybuyers inMarket2 andsold them in Market 3 is the same (assumingno lossesin handling)but that thepriceof the exchangein Market 3 is higher.The middlemanuse thismarginto covertheiroperating,riskand opportunitycosts.

40 15 depends on several things, most nota- the well known demand curve. Figure time picture of bly 1) the state of the fish stocks and the shows an idealized short run the production process; 2) the supply and demand curves for fish. Notice nature of and of fish supplied (usually to middle- that the axes are labeled "Price" quantity curve from the industrial fleet's catch or by- "Quantity" of fish. A single demand men) relationship. catch; 3) the amount demanded by can describe only this partial demand, "A.e. the consumers; 4) the existence of imperfections The other factors affecting 5) the level of infras- preference For fish, the numbers of consu­ in the three markets; the (landing sites, roads, transport sys- mers, the income level of consumers, tructure behind the scene. If tems); and 6) post-harvest losses, price of chicken, etc., are one or more of these increases, for example all meats increased, the new 3.4.3.1 The Industrial Catch if the price of demand curve (labeled New) would lie to the fishery can affect the small- right of (above)the old one. If one or more of The industrial curve scale fishery at many points, helping in these decreases, the new demand some instances and hindering it in others. would lie to the left of (below) the old. The supply of inputs available to the small­ example ice and net- There are many demand curves, of course. scale fishermen, for for all enhanced by the response We can speak ofthe short run demand ting, is frequently a year. Many sellers to the industrial demand. fish, covering a period of, say, of input would be constant in Likewise, roads are constructed, wharves of the shifting factors provided to accom- this case and monthly data wou!d be built, and other services run of the industrial catch. The required. Or we can speak of the long modate the flow teni year two fleets with the fish demand for shark -covering, say, a interaction of the woijld be can be very complicated. The period. Consumer income changes resource case. The nature of the actions of either fleet on the same stocks will important in this available to collected data determines what we can say have an impact on the resource a may be fishing for the about demand. When we actually estimate the other fleet. Both all of the catching some of the other's demand relationship, we employ same stocks, the demand. species as by-catch, or directing their important factors determining target a demand curve graphi­ at different year classes of the same When we represent effort the most important element species, cally, we show of that relationship - that between quan­ Some of the output of the industrial fleet, tity demanded and price. frequently the last day's by-catch, finds its small- One important characteristic of these way into the flow of fresh catch ofthe Ail often find it curves is their steepness or slope. else scale fishermen. Middlemen drops very and less costly to fill their equal, if the demand curve more convenient left to lower right, it indi­ trucks with the by-catch of, for example, rapidly from upper than visit several cates that consumers are relatively insensi­ shrimp vessels rather total amount sites. A proper bioeco- tive to the price of fish: the scattered landing change very much analysis of the small-scale fiahery demanded will not nomic a great reduction in the price. must account for these effects. unless there is If the demand curve is more flat (horizon­ among 3.4.3.2 Consumer Demand tal), there is a gree.ter sensitivity consumers to the price. A measure of this The quantity of fish demanded by consu- sensitivity is called the price elasticity of mers in a period of time is related to its price, demand.17 price of substitutes, the income level of the then a the consumers and the numbers of consu- If the demand for fish is inelastic, as the last three factors short run increase in the supply of fish can mers. In general, to the quantity demanded does as actually result in lower trip revenues increase, only well. The quantity of fish demanded fishermen. For example, if there were as the price increases. This one buyer of fish, and many fishermen with decreases buyer could inverse relationship is a characteristic of good catches on a given day, the

quantity demanded,-Q/Q ito the '7 The price elasticity of demand (P.E.D.) is the ratio of the percentage change in the - "- AP . AQ P percentage change in the price, AP/P; therefore, P.E.D. / -0 -T -AT -0 If the of from 0to 0O and its value is different at every point on a straight line demand curve. The P.E.D. has a range demand is said to be inelastic. P.E.D. is greater than 1,the demand at that point iRsaid to be elastic. If it is less than 1,the At 1it is said to be unitary elastic.

41 Price Supply (new)

"- .Supply (old)

% %

P -

Q3 01 QUANTI Y

Figure 15. Short run supply and dem~and curves. This figure shows the effect of changesin other (thanpriceand quantity)supply anddemand determinantson supply and demandcurves. The shift to anew demand curve can come aboutbecauseof increasesin thepricesubstitutesforfish. A new supply curve might be defined for seasons in which fishing is dangerous. only be induced to increase his purchases The short run supply curve for fish is sim­ through a considerable drop in the price ilar to most other supply curves in its asked by the fishermen. The price, once appearance. This curve, which reflects a agreed upon, would apply to all of the fish positive as opposed to an inverse relation­ sold, not just to the extra catch. ship between prices and quantities offered, can be shifted to the left (or above) by We could likewise draw figures (which are increased costs or a smaller standing stock. not strictly demand curves) which relate A larger stock and/or reduced costs ofeffort some other factor, for example incomes, to will shift the curve to the right (or below) the quantity ofifish demanded. These curves indicating that more will be offered for sale too have elasticities. We can speak of the at every price (Figure 15). The shape of the income elasticity of demand (I.E.D.), for long run supply curve of fish, on the other example, which relates the percentage hand, is influenced directly by the reproduc­ change in the quantity of fish demanded to tive capacity of the resource. A common for­ the percentage change in consumer income mulation of this curve incorporates the levels. Schaefer-type equilibrium yield curve (Fig­ Since each species or commercial class has ure 16). its own demand relationship, it is reward- If we compare the price elasticities ofsupply ing to compare the impact of the same (commonly called the price response) for demand determinant for different species or two short run curves we would expect the classes. For example, the income elasticities elasticities to be greater for that situation in of demand for different commercial classes which there was a larger standing stock. In both casesfishermenrespondtopriceincen­ often differ greatly. Research in Central - increasing the that as the income tives by fishing more America demonstrated would expect the suc­ changed, their demand for supply of effort. We of consumers of effort We ou d eexp loi­ more desirable (and expensive) species s for less desira- cess of that effort applied to an underexploi­ increased while the demand than if it were applied decreased. The more desira- ted stock to be greater ble fish actually with more ble fish were caught by fishermen capital intensive operations and the less In the long run where all inputs to fishing, desirable fish by the poorest fishermen including capital (vessels) are allowed to (Coslit, Lampe and Sutinen, 1980). vary, the long run price elasticity of supply capacity decides how much fish to pur- is constrained by the reproductive An individual measure is greater (more by considering its cost and the cost of of the stock. This chase equilibrium points before MSY, other things he needs or wants. He is usu- elastic) at as much as he decreasing as it approaches the maximum ally restricted from buying very inelastic he wants by his yield. At points near MSY it is wants of everything MSY: as price incen­ usual approach to estimating and is negative beyond income. The and induce more effort, relations is not to analye all of tives create profits thesedemand decision processes for all c~onsimers, decreases.the total output of the industry actually but to examine the transactions that actu­ ally take place in the market together with 3.4.3.4 Bottlenecks, Power and Losses some of the other important factors in indi­ vidual decisions. These transactions reflect The three remaining conditions mentioned the results of all individual decision pro- above - the ievel of infrastructure, the cesses grouped together. existence of market imperfections and post­ harvest losses - can further reduce the 3.4.3.3 Supply quantity of fish reaching the consumer. These three conditions can be mutually Supply relationships indicate the quantity reinforcing as well. Conditions such as of goods that producers are willing to sell unreliable transport or a lack of roads to given the price of the goods, the cost of pro- scattered landing sites can cause the waste ducing the goods and, in the case of the of landed fish. Such conditions also contrib­ fishery, the state of the resource. As with ute to the ability to exercise market power ­ demand, supply relationships measure a power which can cause the wastage of fish. flow of goods in a time period. We can speak of the supply of all fish in a given year, the Undervalued fish is either discarded at sea supply of certain species or classes, the or at the landing site. The undervaluation supply by all firms (industry) or by certain problem is most often a question of preferen­ types of firms, for example all gill net ces related to taste, appearance, boniness, fishermen. taboos, etc. As such it requires the attention

43 Price

MSY

QUANTITY

Figure 16. Long run supply and demand. The long run supply curve(afterCopes, 1972) incorpo­ ratesthe long run responseof the stocks to increasedfishingeffort. The intersectionof demand curve D2 with the supply curve occurs in a situationof biologicaloverfishing.

44 of people speciaJizing in food habits, fish (inputs) and monopsonistic buying power processing and marketing - in the sense of (fresh fish) in the harvest sector, it is more advertising and education. Fish which is likely that 1) the quantity flowing through valued can be wasted, as well, as it deterio- the system will be less than that which in a physically and economically inef- would flow through under competitive con­ rates will be ficient delivery ststem. ditions, 2) the price to the consumers higher than that determined by competi­ The consumer demand for fish,is influenced tion, and 3) the total income 'o fishermen by the quality of the fish supplied sfnd. to will be smaller. Furthermore, the advantage some extent, by the consistency of that that the monopolist has over buyers or the supply. In an attempt to assure a consistent monopsonist has over sellers tends to be supply of fish, arrangements are often made greater as the respective demand or supply back through the system of retailers, whole- curve is less elastic. salers, middlemen and fishermen. These arrangements can appear as the vertical Groups of competing fishermen,, facing integration of firms in which, for example, either or both monopoly sellers of inputs middlemen acquire boats, transport and such as ice or monopsonist buyers of their retailing outlets. The middleman thereby fresh fish, often organize to develop their becomes a complete delivery subsystem own market power. They form fishermen's unto himself. Or these arrangements can organizations or cooperatives so that they appear as mutually beneficial (initially at can buy and sell as a unit. These organiza­ least) agreements between, for example, a tions further develop, at times, by integrat­ middleman and a set of fishermen. In return ing marketing, processing, transport and for their guaranteed supply (periodic retailing activities. the fishermen may though it may be) over time the for gear or for personal debts It is frequently observed that receive loans in a particular 1979). In this regard the middleman numbers of buyers or sellers (Smith, Those remaining may be represented in both the input (to fish- market decreases. can scale of operation (size of ing) market and the fresh fish output have come upon a a smaller cost per-unit out­ market. plant) that has put. For example, larger more capital inten­ The The three markets we have discussed can sive plants may have this advantage. "perfectly" - maximizing the flow flow of fish may be such that one or two operate supply. of fish through the system at minimum cost plants can adequately handle the of these arrangements as long as However, 83 the numbers of buyers or - in spite to there is no exercise of market power, i.e. as sellers in a market decrease, the ability as no one exploits either buyers or collude and to set prices increases. This con­ long by bar­ sellers. The opportunity to exercise market centration of buyers or sellers, aided exists when one or a few individuals riers to competition, can easily increase its power of can influence prices because they are pro. profits beyond what the same level tected from competition by another set of investment would make in another barriers. These too may be physical barriers industry. to isolation (infrastructure), informa­ due buyer of fish can increase barriers (on prices) or economic barri- The monopsonist tion offering fishermen a price ers (high capital costs, risk, etc.). his profits by below that which would be offered if compe­ power Control over the demand for a good by one tition existed. The exercise of this few people is termed respectively mo- may even result in resource conservation. In or a col­ nopsony cr oligopsony power. Market this instance the fish buyer is in effect power in the hands of a single or of a few lecting the resource rent. suppliers of goods is termed, respectively, power. In our simpli- The degree of market power is sometimes monopoly or oligopoly between the market system we would expect equated with the difference fled three and that charged for there are many fishermen and many price paid for inputs that difference, however, includes consumers. There exists the possibility, output. This power all of the costs incurred by, in this example, then, of three distortions: monopoly are monopsony power in the middleman. Frequently these costs in supplying inputs, casual observer. fresh fish, and monopoly power in underestimated by the buying used indicator of the selling processed fish. Another commonly exercise of market power is the percentage extent that opportunities exist for of the final retail price that fishermen To the opportuni­ exercising both monopolistic selling power receive for their fish. While the

45 ties for the existence of market power can be calculated from them also require what is easily be identified, only careful analysis of called time series data. There is no substi­ specific cases can reveal exploitation (see tute for collecting this through time if any­ for example, Epler, et al 1980; Scheid, et al thing is to be said about supply or demand 1980; and Coslit, Lampe and Sutinen, 1980). curves. 3.4.4. Information Requirements The particular need for other information from outside the harvest sector is dictated The review of the fishery system above by what one wishes to know and what inter­ should have demonstrated the need for a ventions are planned. For example, if back­ variety of economic (bioeconomic) informa- ground information and observations tion in understanding the forces at work in suggest distortions in the supply of inputs, the system. We have attempted to show that an analysis of the structure and perfor­ the interaction of biological, cultural and mance of input markets may be called for. economic forces complicate intended The important time series of transactions changes. Furthermore, the goals of assist- can be coupled with, for example, produc­ ing fishermen and consumers, of conserv- tion information to identify the supply and ing the resource and of rationalizing the demand relations in that market. harvest sector can be frequently incompati­ ble ifnot pursued simultaneously. We exam- In the primary market for fresh fish the time ined the fishing process and saw that the series of transactions can be coupled with production of fish was affected by the legal information on other determinants of and economic milieu in which it operated; demand by the middleman (his costs and that the potential supply from this sector selling prices) in order to identify this rela­ could be affected by the nature of the inter- tionship. Furthermore, to the extent that vening markets and by a number of combi- there is competition in the purchase offresh nations of supply and demand elasticities, fish by middlemen - or perhaps a lack of We have reviewed the operation of this sim- counterindications - the demand by the plified, yet complete system so that the rela- middlemen may be interpreted to reflect tionship between the harvest sector and the demand by the consumer. A constant per rest of the system would be clear. The infor- unit margin for the middleman is assumed mation requirements for economic analyses in this case. depend upon what is being examined. On the other hand, all of these analyses draw If commercial fleet catch or by-catch is com­ upon a set of data whose elements are peting in the same markets as small-scale related to each other by the decisions of catches, then an equivalent time series of fishermen. Therefore, rather than repeat prices and quantities for this industrial out­ even the limited uses of economic data pre- put is required (see 4.3). fiedsented as inthat this necessary chapter, informationto study supply is classi­ and In sum then, the harvest sector can yield demand relations and that necessary to crucial information for the determination of examine productive processes, supply and demand Transaction in all three markets. information by itself can yield 3.4.4.1 Transactions interesting statistics but must be comple­ mented with information from outside what The data identified for collection in the next we have strictly defined as the harvest sec­ chapter will include that on transactions in tor in order to reveal supply or demand the first and second markets: on the pur- relations. chase ofinputs and the sales ofoutput (fresh fish). Transaction information does not 3.4.4.2 Productive Processes represent either demand or supply curves, merely their intersections through time. We have identified the incentives to enter or When combined with additional informa- leave the fishery as, respectively, industry­ tion on other supply and demand determi- wide profits or rents which offer the oppor­ nants from both inside (the harvest sector) tunity for "above normal" returns, and and from sources outside (such as price returns which are less than those which indexes or per-capita income estimates), could be earned in the ncxt best use of one's these transactions can be made to reveal the labor and money. In the open access fishery supply and demand relations. we expect these rents to be zero once the fishery is in equilibrium regardless of the The supply and demand relations and level at which the stocks are exploited. many of the important elasticities that can Earning the normal return requires the

46 fisherman to make many decisions concern­ ing the physical process of fishing and its economic consequences. The physical production relationship requires information on the kinds and quan­ tities of inputs used in a period of time. In the short run the physical process involves combining fixed inputs such as the hull and gear with variable inputs like ice, bait and crew services. Each firm takes these inputs and finds the different levels of output resulting from varying the input combina­ tions. Over the long run even the fixed inputs are changed. Changes in the physical production process are motivated by the different returns each mix of variable and fixed inputs can bring. The returns are determined by the quanti­ ties and prices of inputs and the quantities and prices of the output they produce. Varia­ tions in the prices of outputs or inputs bring both short run and long run adjustments in the production process. If we add to information on the prices of inputs to the physical production relation, we can generate a total cost of production relationship. If we add information on pri­ ces of output to the production relation, we can arrive at a total revenue relationship. Combining these cost and revenue relation­ ships yields a profit or net revenue relationship. In sum, we will see that most of the impor­ tant fishery-wide economic information we seek (andwhich can be obtained in the harv­ est sector) is generated from data which des­ cribes the fishing activity of individual firms. Supply and demand information and that on the economic and physical aspects of production is derived from data on the technical and economic decisions made by individual fishermen and on the impact these decisions have on the resource.

47 Chapter IV DATA REQUIREMENTS

are available for 4.1 BIOLOGICAL/ ECOLOGICAL until more specific data DATA individual species. Grounds 4.1.1 Habitat/Ecosystem Data 4.1.1.2 Extent of Fishing has 4.1.1.1 Biological/Ecological Once some kind of a resource inventory been compiled, it is advisable to conduct an which are An inventory ofthe species currently exploi- inventory of fishing grounds (and species which might presently supporting the fishery or which ted by the fishery effort. For in the future) is an essential could support expanded fishing be exploited known to in designing data collection proce- example, for demersal species first step in depths up to 50 which will provide information neces- inhabit coastal waters dures accurate estimate of sary for management and development meters, a reasonably should include all the extent of harvestable fishing grounds purposes. This inventory area of the life histories of the could be obtained by calculating the that is known about within the 50 exploited species and the ecologi- offshore continental shelf principal distinctions might be cal factors which affect their distribution, meter contour. Finer Some made between different types of substrate relative abundance and availability. will items in the inventory (mud, rock, sand). This inventory of the important fishing gear Ee longevity, size (age) at first matur- require some knowledge of the should species or groups ity, miaximum size, fecundity, temperature used to exploit individual tolerances, feeding habits, des- of apecies. Rough estimates of expected and salinity an expansion cription of habitat, migrations, spawning increases in catch following new grounds can be seasons and locations, growth rotes, and of the fishery to in size, morphol(,gy and obtained by multiplying catch per unit area sexual differences fisheries which behavior, data for existing small-scale use certain gear types and exploit certain Once these data are obtained, it should be species combinations by the area of the new possible to group individual species which grounds which could be exploited ifexisting share common ecological and biological fishing gear or practices changed. and which are likely to be characteristics and Ecosys­ harvested together. Examples of such 4.1.1.3 F 'oduction Rates groups are small pelagic schoolirg fish tem Dynamics caught in purse-seines, large non-migratory prim­ line or It is important to know the sources of demersal fish caught with hook and exploited lobster which are ary production which support several species of spiny primary production Such groupings can stocks and whether harvested in traps. sources is high, low or for a crude definition of a unit from these various form the basis Sources of primary production stock for resource assessment purposes moderate.'

plants, a process which converts Primary production is defined as the photosynthetic production of organic matter by I supports the entire ecosystem. The sunlight into plant tissue which is available to herbivorous animals and therefore "trophic groups" at different "levels," i.e. phytoplankton (microscopic plant life) at the ecosystem is composed of different fish at level, zouplankton (tiny herbivorous animals) at the second trophic level, small plankton-feeding primary, trophic refers to the production of the third level, etc. until ultimately reaching top predators such as man. Secondary production animal tissue through growth and reproduction.

49 Previous Page B1 in tropical marine ecosystems are 1)phyto- water habitats where yields have been plankton in coastal oceanic or estuarine sur- empirically related to an index which corn­ face Naters, 2)benthic algae, 3)coral reefs, bines the quantity of total dissolved solids 4) rooted aquatic plants, or 5) detritus and the depth of lakes (Henderson et al., derived from mangrove swamps. If average 1973; Ryder et al., 1974; Ryder, 1978). A sim­ annual rates of primary production are ilar exercise has been conducted with large known and a reasonably accurate rate of river systems by relating yields to the size of energy transfer between succeeding trophic drainage systems (Welcomme, 1979). Exten­ levels can be assumed (usually 10%), it sions of this procedure tothe coastal marine should be possible to estimate theoretical habitat would require a thorough catalogu­ rates of secondary production for certain ing of physical and chemical parameters exploitable components of the ecosystem. and a search for possible mathematical Some proportion of this production, i.e. the relationships with harvest data. energy which is used for growth and repro­ duction of offspring which survive to reach 4.1.2 Stock Assessment Data a size (age) where they can be exploited by the fishery, is converted into biomass. Sur- 4.1.2.1 Introduction plus biomass (i.e. biomass which remains Although a number of approaches to after natural mortality has been accounted resource assessment were discussed in for) can be harvested. Chapter III, further discussion of resource will focus on data In order to assign individual species to spe- assessment in this guide compo- requirements and collection methods for the cific trophic levels or ecosystem per habits data are needed. surplus production model and the yield nents, feeding it is appropriate, a dis­ Furthermore, if predator-prey relationships recruit model. When these data, it might be tinction will be made between preliminary can be deduced from more detailed assesse­ possible to predict the effect on total stock assessments and removing ments. Other types of resource evaluation biomass and yield of selectively are less (by fishing). Other types of are not excluded because they certain species because the assessments per­ competition between species (such as for the important, but result in formed in developing countries should be same food source or space) can also which estimates for individual based on available data and/or data unreliable yield easily obtained. stocks which are treated as separate biologi- can be most cal entities. The most common types ofassessment data Data which exist in developing countries are 4.1.1.4 Hydrographic catch and effort data. Estimation of growth Estimates of the amount of freshwater flow and mortality rates for use in more sophisti­ into major estuaries and coastal waters and cated models can sometimes be accom­ information on coastal currents, tidal flow plished by fairly routine compilation of size and amplitude, and offshore upwellings are frequency data from samples of commercial important for identifying factors which landings or from research techniques such limit rates of primary and secondary pro- as tag and recapture studies. One purpose of duction, principally by 1..niting the availa- this guide is to focus on relatively low tech­ bility of nutrients necessary for nology research techniques and data collec­ plant growth. Hydro- cion methods since the necessary tools photosynthesis and and graphic data may also contribute to an (vessels, equipment and computers) migratory patterns of skills (personnel trained in fisheries understanding of the may be juvenile and adult fish, the dispersal ofeggs science, mathematics and statistics) and larvae and the spatial distribution of lacking. fishing effort. 4.1.2.2 Catch and Effort Data 4.1.1.5 Other Physical/Chemical Data Catch and effort statistics are extremely These data are important since they provide useful for resource assessment purposes. As information concerning the environmental outlined in section 3.2.4, these assessments limits to production and habitat informa- may be preliminary and involve simply an tion for resource inventories. Furthermore, evaluation of temporal trends in catch, physical/chemical data can be related effort and catch per unit effort (CPUE) data. empirically to catch statistics and thus form At this level, useful conclusions can be the basis for yield predictions. This tech- reached from several years of data. Qualita­ nique has been successfully applied to fresh- tive judgements can be made about the

50 degree of resource depletion and the proba- most small-scale food fisheries, the quantity ble outcome of increased or reduced fishing of fish which is discarded at sea is not sig­ effort, i.e. "will catch increase or decrease nificant. Losses due to spoilage, however, and what will be the impact on the can be substantial when proper storage and resource?" More thorough quantitative handling procedures are not followed assessments incorporate catch and effort aboard the fishing vessels. Furthermore, data as parameter estimates in when catches are low, fishermen are more mathematically-derived models and require likely to sell a larger proportion of their a longer and a more reliable set of annual catch directly to retailers or consumers in catch ard effort statistics. order to increase their income. Both preliminary and more detailed assess- Fishing effort is frequently estimated either ments may be performed with catch and as the number of operating units (vessels, effort data for a combination of species or gear, fishermen) in use during a certain stocks which are exploited by the same gear period of time or some measure of the time in the same general location or for individ- spent fishing such as the number of trips, ual species or unit stocks. Analyses per- days, or hours. More useful estimates o' formed with combined data sets, however, effort include both a time component and an may provide more reliable maximum yield estimate of the relative capture efficiency of estimates since they account for species different gear and vessel combinations. A interactions (for example, predator-prey net which measures 50 x 100 m and is fished relations) which are not accounted for in for six hours, for example, should catch unit stock applications which ignore the approximately twice as much fish as a 25 x biological and ecological relationships 100 m net which is fished for the same between species. period of time, assuming each net is identi­ cal (same mesh size, material and construc­ Catch and effort data which are required for tion) and is being used by equally skilled surplus production analysis are generated fishermen under similar conditions. The in two stages. Primary data are collected as best estimates of the amount of time actu­ fish are landed and delivered for sale to ally devoted to catching a certain quantity primary dealers. In the second stage these of fish are defined by the nature of the data are compiled according to specific time fishery. For a handline fishery, for example, periods (months, years), gear types and gen- the time that the hooks are in the water eral fishing locations. Effort data should would be most reliable whereas for a pelagic account for the capture efficiency of the gear purse seine fishery it could be the time spent which is used as well as the amount of time searching for schools of fish. Guidelines for spent fishing. Adjustments are required in estimating fishing effort for five selected order to correct landings statistics for fish small-scale fishing gears are presented in which is captured, but not landed. Table 1. For further advice on how to esti­ for different types of fishing data are available for indi- mate effort Once sufficient are compiled into a vidual time periods, they time series of annual c. tch and effort esti- Standardization procedures are aimed at mates which ideally should represent a correcting for changes in fishing power period of at least five to ten years in the which take place over time as improve­ historical de,;elopment of the fishery and ments in gear and vessel technology include yearn; of high and low exploitation, increase capture efficiency. Thus, if 100 Time series catch and effort statistics are hours of fishing in 1970 was twice as effec­ analyzed to generate estimates of maxi- tive, because of improvements in gear mum sustainable yield and the correspond- design, as 100 hours of fishing in 1960, ing amount of fishing effort. standardized effort in 1970 would be double the amount of effort in the reference year Catch data are usually estimated from the (1960) even though the number of hours weight (or, in some cases, volume or even fished remains the same. Effort standardi­ numbers) of fish which is sold to primary zation procedures have not generally been dealers plus estimates of how much fish is applied to small-scale fisheries, mostly caught, but not sold, to primary dealers. because historical changes in capture effi­ Sources of post-harvest loss include fish ciency have been minimal or difficult to which spoil and fish which are discarded at quantify. sea, gutted and either sold, given away or kept by the fishermen before they reach the A problem which is often ignored is the cap­ primary dealer (see Brander, 1975). For ture of species which are harvested inciden­

51 Table 1: Some factors affecting the performance of five selected small-scale fishing gears and appropriate definitions of fishing effort and catch per unit effort for each one. EFFORT DEFINITION CATCH PER GEAR PERFORMANCE FACTORS (time x power) UNIT EFFORT Gill net Mode of fishing (surface or bottom; Time fishing' x area of net Catch/hr/standard drifting or anchored), type of twine, net area mesh size and shape, area of net Handlines Type of bait (or lure), hook size, Time fishing' x number of Catch/hr/fisherman, mode (trolling, bottom, by hand or fishermen, lines or hooks line or hook mechanical reel), number of lines or hooks Longlines Bait, hook size, mode (surface, mid- Time fishing' x number of Catch/hr/standard water or bottom), number of hooks hooks number of hooks Fish pots Volume, bait (if used), mesh size Submersion time x number Catch/pot/day or opening between laths, design, of pots hauled entrance size Beach seine Mesh size, length of net Number of hauls Catch/haul

I Estimated as the time the gear was actually in the water or the time spent on the fishing grounds tally to the predominant species. The catch minimum size (or age) at which fish are first of certain species bears little or no relation captured through use of dynamic pool mod­ to the fishing effort expended during a els (see section 3.2.4.2 (b)) require a consider­ given trip if they wvere not the primary spe- ably more complex set of data than surplus cies which were being sought. Catch and production models which are applied to a effort data for incidental species should not time series of catch and effort data. be compiled for assessment purposes. An Although the data required for dynamic acceptable procciure for eliminating these pool analysis are more difficult to obtain, species is to ,only compile catch and effort they can be collected during a fairly short data for species which make up more than period of time (perhaps one or two years); no some minimum percentage (say 10-20%) of time series is required. the catch during a given fishing trip. An additional problem is posed when a sin­ gle unit stock assessment is performed Parameter estimates which are required for using effort data which are compiled from dynamic pool analysis refer to exploited more than one gear type. This is a particu- populations or unit stocks, but they are larly relevant problem for small-scale obtained from observations of individual fisheries which harvest riany species using fish which are caught by fishermen or from a variety of gear types and for stocks which fish which are collected for research pur­ are exploited simultaneously by small-scale poses or during exploratory fishing surveys. and industrial fisheries. Stock assessments Some of the necessary parameters are listed which rely on combined effort estimates in Table 2. They are generally of two types: require comparative fitld studies of the 1) those which refer to the rate at which catch rates of different gear types used to some biological process (birth, death, exploit the same stocks during the same growth) takes place and 2) those which refer period of time. A description of techniques to ages (or sizes) when these processes begin might be used in such studies is or end. Because the processes have a biologi­ which origin, the data are referred to as "vital" beyond the scope of this guide. In the cal absence of this information, unit stock statistics. assessments are carried out for individual unit fisheries, i.e. fisheries which are char­ acterized by a single gear type and fishing Although it is not the purpose of this guide location, to describe data analysis techniques, some of the more common conceptual models and 4.1.2.3 Vital Statistics analytical procedures will be mentioned and briefly described in Chapter V since Quantitative assessments which generate they are used to convert observations into predictions of maximum sustainable yield more refined data which are in turn ana­ for a given amount of fishing mortality and lyzed IDproduce information useful for the

52 *Table 2: Definitions of key parameters used in dynamic pool models and data required to estimate them; data requirements are limited to methods discussed in this guide. PARAMETER DEFINITION DATA Growth rate (K) Rate of increase in size per unit time, Size frequencies of catch samples by gear type expressed as a time-dependent coef- and location over time; sizes of individual fish ficient without units at tagging and recapture and dates of release and recapture; size, distances from scale (otolith) focus to annual radii and scale edge for individual fish

Maximum limiting Theore 'cnl !imiting size attained Same as above or can be assumed to equal maximum size (L. , W ) by an average individual in the pop- observed size when exploitation is not severe ulation, expressed in terms of length or weight to Theoretical age at which growth begins Same as above or can be assumed to equal zero Total mortality Instantaneous rate at which numbers Relative abundance (CPUE) of a given age)group rate (Z) in the population decline pcr unit over time or relative abundance of two more time for any reason successive age groups at any point in time; number of tagged fish recaptufed per unit effort in successive time intervals

Natural mortality Instantaneous rate at which numbers Relative abundance of an age group over time or rate (M) in the population decline per unit relative abundance of two or more successive age time as a result of predation, groups at any point in time for an unexploited disease, etc. population;annual changes in Z and fishing effort; by subtraction if Z and F are known

Fishing mortality Instantaneous rate at which numbers Number of tagged fish recaptured per unit effort rate (F) in the population decline per unit in successive time intervals plus total number time as a result of fishing tagged initially or by subtraction if Z and M are known

Recruitment (R) Number of recruits (juvenile fish) Relative size of recruiting ege classes can be entering the exploitable population estimated from repeated survey data (CPUE) or per unit time catch rates in a fishery which harvest pre-recruits

tr, Ir Age (length) at recruitment Age (length) composition of commercial catches t.,1c Age (length) at first capture

management of fishery resources." Parame- order to produce actual biomass yield ter estimates, by themselves, are of no use estimates. for management purposes; they simply pro­ vide data necessary for assessments. Furth- 4.1.2.4 Exploratory Fishing Surveys ermore, assessments will not be feasible unless a minimum set of parameter esti- Biomass estimates of exploited or unexplo­ mates isavailable. As mentioned in section ited stocks can be based on catch rates 3.2.4.2(b), not all of the statistics listed in obtained during exploratory fishing sur­ Table 2 are necessary in every case. Modi- veys. Mean catch rates are calculated from fled versions of the original model only repeated sampling with the same gear fol­ require, for example, estimates of mortality- lowing the same fishing procedures and are /growth ratios and size instead of age- converted into catch per unit area estimates specific estimates. which are in turn multiplied by the area which the stock inhabits to produce bio­ Once all the necessary parameters are esti- mass estimates. Surveys can be used to esti­ mated, maximum yield can be predicted as a mate biomass only if catch rates can be function of fishing mortality and the age or calculated on a per unit irea basis and if size at first capture from yield tables such as some reasonable estimate of the capture those compiled by Beverton and Holt (1966). efficiency of the gear is available. For best Since recruitment is frequently unknown, results, therefore, some preliminary data yield estimates are given on a relative per- concerning gear selectivity and capture effi­ recruit basis and must be multiplied by the ciency for individual species and sizes of number of recruits which enter the exploita- fish are necessary before it can be assumed ble stock during a particular time interval in that the catch rate of the gear which is used

2Amorecom. stediscussion of methods used for compiling vital statistics is given by Gulland (1969) in a manual which is available in both English ane. Spanish versions. Another very useful reference is in the handbook by Ricker (1975).

53 is directly proportional to the abundance of ent groups. Development projects may fail tIie exploited population in some defined in cases where tension exists between airea. In practice, exploratory fishing sur- groups (Cochranc, 1979). Tensions are veys are carried out in the absence of such manifested by various types of behavior information; in many cases, they are ranging from discrimination all the way to designed to answer more basic questions of open warfare. The degree of tension should interest to fishermen such as what sizes and be evaluated by examining intergroup atti­ species of fish are available, where and tudes, beliefs, and values and by describing when they can be harvested and in what overt manifestations ofthese tensions, such quantities. Biomass estimates are usually a as employment discrimination or economic secondary objective of such surveys, subversion. 4.2 SOCIOCULTURAL DATA 4.2.2.3 Numbers and Locations of Potential Project Participants 4.2.1 Introduction In the planning stages of a development pro­ Specific types of sociocuitural information ject where it is uncertain whether new tech­ were identified in section 3.3 which describe nology will be introduced or the numbers of some of the general elements of the harvest- fishermen increased (or both), it is essential ing sector of small-scale fisheries and which to determine numbers of (a) active fisher­ are also important factors to be considered men, (b) unemployed fishermen, and (c) in making appropriate management or unemployed or underemployed non­ development decisions. In this section, the fishermen who could potentially become types of data necessary for generating this fishermen. Data should be prepared which information are described, indicate where individuals who belong in three categories are located. General Background Information these 4.2.2 4.2.2.4 Traditional Communication Channels noted in the preceding chapter, general As needed in order background information is to understand how the fishery functions Data concerning traditional communica­ and to evaluate the effects of intervention in tion channels will consist of a listing and the fishery. The basic sociocultural informa- distribution of available media (newspa­ tion generated is such that it will be of use to pers, magazines, radio, television, cinema) biologists and economists as well as to and locations where information is dissemi­ fishery administrators. The following dis- nated (such as meeting halls, public square, cussion describes data needed for specific market place, school house, the mayor's information types. door). The language used in communica­ tions should also be noted. This is especially 4.2.2.1 Identification of Groups important in multilingual communities have different be assumed that all small-scale where different languages It cannot Basic data on infor­ nin a given country or region are statuses and functions. fishermeni a gimes ty eon ar- mation networks should also be delineated. identical. Sometimes they belong to differ- This includes identification of key individu. ent tribes or ethnic groups, sometimes they als who are trusted information sources for practice different religions, sometimes they networks of individuals. Attitudes and beliefs which relate to the various communi­ snmetimesmanifest differentdifferent politicalgroups areloyalties, defined and by cation channels and languages should be of fishing (such determined in order to identify the most their technology and style ver­ versus line fishermen or inshore as net for disseminating pro­ offshore). In some cases, the position of appropriate means sus ject information. the fishermen as a group distinct from other groups will be important. The key factor is that the groups identified often act on the 4.2.2.5 Level of Community Develop­ basis of their own perceptions of their differ- ment ences from other groups. The data which are A listing of available services, many of needed include location, numbers, and spe- which are also needed to evaluate infras­ cific identifying characteristics for each tructure, is required in order to determine group. the level of community development. Servi­

Tension ces include schools, government offices, util­ 4.2.2.2 Degrees of Intergroup ities (water, electricity), banks, It is not enough to simply distinguish differ- transportation facilities (trains, buses,

54 lines), mass media, medical is made to intervene in the fishery for devel­ roads, shipping it is services, and wholesale and retail outlets. opment or management purposes, necessary to evaluate the social structure of 4.2.2.6 Occupational Structure the occupation of fishing. These data will be social impact of Occupational structure is determined by useful for determining the proposed changes as well as facilitating examining data which summarize (a) types their communication to the target of occupations which are available; (b) dis- population. tribution of workers in the various occupa­ tions; (c) skills required; and (d) degree of 4.2.3.1 Fishing Gear Types and Owner­ demand, underemployment and unemploy- ship Patterns ment for each occupation. Data on fishing gear types should include 4.2.2.7 Numbers of Small-Scale local names for all types ofequipment(such Fishermen as boats, nets, lines, hooks, harpoons, motors, and sails) along with descriptions equipment are Numbers of employed and unempoyed of where different types of small-scale fishermen are considered above used, how they are used anA by whom. A (4.2.2.3 and 4.2.2.6). description of ownership patterns should an evaluation of the relative impor­ include ownership (i.e. 4.2.2.8 Availability of Alternative tance of different types of for Fishermen individual versus joint ownership, lease, or Occupations patterns This information can be provided with data rent). Descriptions of ownership which describe the skills which are required should be prepared for individual regions and the demand for and gear types. Procedures for transferring for other ccupations In workers in those occupations (4.2.2.6). ownership should also be investigated. some cases, it may also be important to 4.2.2.9 Temporal Distribution of Fish- know how many owners operate their own ing Effort equipmen't. The amount of fishing effort usually varies 4.2.3.2 Crew Size and Social throughout the year. Data required include Composition days of the week, and the time of day, and social compo­ (or seasons) of the year which are A description of crew size months a classification which devoted to fishing and the times when fish- sition begins with Ifthe variation shows the relationships between the posi­ ing is more or less intensive. (cap­ on the gear types or fish- tions held by different crew members in effort depends on each type which are used, data should tain, cook, net man, engineer) ing practices of each type of person also be recorded for specific types of fisher- of vessel. The number for changes in effort, aboard each type of vessel should be deter­ ies. Finally, reasons the of fishing throughout the year mined along with data summarizing gear, or style crew members. should be determined. Explanations would kinship links between such as the weather and include factors Criteria for'Crew Selection holidays. 4.2.3.3 collected as a part of 4.2.2.6 will pro­ 4.2.2.10 Local Knowledge About Fish- Data vide some information which reveals skills ing and Fish required of different crew members. Addi­ Primary data needed here include a descrip- tional information is needed to determine actually used in crew selec­ tion of principal gear types and the ways in what criteria are Local names for tion. It is important to learn the relative which they are employed. and/or should he recorded along with importance of kinship, friendship, all gear types in the selection of the common names of all fish known to the occupational efficiency have commer- crew members. It is also essential to deter­ local fishermen. Fish which as value should bedes- mine whether or not other criteria such cial and/or subsistence are used as crite­ least in terms of their habitat, ethnic group membership cribed at of fishing crews. ecology, behavior, and relative availability ria for selection as perceived by the local fishermen. 4.2.3.4 Degree of Occupational 4.2.3 Social Structure of the Mobility Occupation It is important to determine whether or not classes of fishermen have an Once basic data are collected and a decision the different

55 opportunity to improve their position. For accord-Ing to the various social categories example, is it possible for a deck hand to defined in sections 4.2.2.2, 4.2.2.6 and own a vessel of his own some day? Do hard 4.2.3.2. workers obtain better positions, more income, or the opportunity to borrow funds Power is defined as the ability of one person to become owners? The degree of occupa- or group to influence another's behavior. tional mobility should be determined for Data requirements include a description of various fishing types, equipment classes, both the official and unofficial political and distinct geographical regions. organizations which exist at the local level, as well as an identification of opinion lead­ 4.2,3.5 Interaction Between Crew ers who klfluence the behavior of fisher­ Members and Between Owner and men. As applied specifically to fishermen, Crew data requirements also include a descrip­ tion of formal government power groups, The social relations (such as relative status the services they provide to the fishermen, and the degree of power one individual has and the local attitudes towards these servi­ over another) which exist between the crew ces and government personnel. They also members (4.2.3.2) and between the crew and include formal laws governing the use of the the owner should be determined. A descrip- ocean by fishermen, as well as informal tion of these interactions should include agreements affecting who fishes when and time spent at sea and on shore. where. 4.2.3.6 Degree of On and Off Vessel Occupational Specialization 4.2.4 Innovation Data Needs This information consists first of the job For purpose of the data collection scheme description - both on and off vessel - of presented here, an innovation is defined as each vessel position (4.2.3.2). Additionally, a new object or idea; thus it includes such all shoreside occupations directly related to things as a new piece of equipment, a new the capture fishery should be abstracted way to use an old type of equipment, a from data collected as part of 4.2.2.6 (for fishery management plan, or an institution example, middlemen, boat builders, out- such as a fishermen's cooperative. The data board engine mechanics), and a more included in this category serve to determine detailed description of the skills required for factors which can either facilitate or impede these jobs, as well as an estimate ofthe time the introduction of an innovation. each worker devotes to the fishery, jhould be provided. 4.2.4.1 History of Innovative Behavior 4.2.3.7 Relationship of Fishing Groups with other Social Groups in Society It is helpful to understand the events sur­ rounding recent attempts to introduce Data are required which describe the fisher- changes which have affected fishing com­ men's social relationships with other social munities. Important data include historical and occupational groups in the society. sketches of the types of innovations which Relationships such as group membership, were attempted, the way they were intro­ cooperation, noninvolvement and hostility duced (i.e. by whom, who were the first need to be identified (groups are identified users, how long did it take until most fisher­ as part of 4.2.2.1). Types of relationships men used it, why did most fishermen adopt which represent different small-scale fisher- - or not adopt - it?), including descriptions ies and geographical regions should be of failures and explanations of why certain defined. Important distinctions may exist, innovations failed. for example, between urban and rural areas. 4.2.3.8 Local Distribution of Wealth 4.2.4.2 Attitudes Towards Risk, and Power Change, and Investment The primary data needed here first of all Data requirements includo descriptions of require a actermination of what is locally variations in attitudes towaird risk, change, considered as wealth. Is it money, land, cat- and investment in the fishery. For example, tie, many children, boats, many followers or under what conditions would fishermen try friends, or some combination of items? Once a new gear type or invest in a more expen­ the definition of wealth is known, the local sive vessel? Would they make such changes distribution of wealth must be determined today, and why or why not?

56 4.2.5 Occupational Preference and changes would have on the lifestyle of the Training Data Needs people who are affected. A great deal of research has clearly demon- 4.2.5.4 Perceived Innovation strated that a person's occupation and atti- Complexity towards the occupation play a large tudes or not role in self-perception, health and social It is important to determine whether Job satisfaction has been fishermen who are intended to benefit from adjustment. the ncces­ related to a wide range of social, psychologi- the innovation believe they have health-related variables sary skills to implement the necessary cal, economic, and in from level of productivity, job turn- changes or can learn them. Variations ranging types, geo­ over, family violence, and psychosomatic these perceptions across fishing to longevity (HEW, 1973; Gelles, graphical regions, and in relation to indi­ illnesses age, education, 1974; Pollnac & Poggie, 1979). Thus, one vidual attributes such as cannot assume that non-fishermen will will- and status must be determined. ingly become fishermen, or vice versa, and/or that they will adjust satisfactorily 4.2.5.5 Perceived Innovation after such a change. It is therefore essential Compatibility to understand what determines job satisfac- It is important to determine whether fisher­ tion and job pre' arences in situations where men feel that they will be physically and occupational changes are anticipated. mentally comfortable adopting a particular innovation. Will the use of the innovation 4.2.5.1 Attitudes Towards Fishing by affect valued social relationships (such as Non-fishermen relationships between owners and crew, members and between fisher­ changes require addition of between crew fishermenIf proposed to the workforce, it will be neces- men and the middlemen), and if so, can problems be resolved? Will the tem­ sary to determine attitudes towards fishing these workers. poral demands of the innovation (night among the pool of available or long trips of relationships betweea fishing instead of day fishing Secondly, analysis fit their present lifestyle, groups and other groups in the versus short trips) fishing possible to adjust the innova­ (see 4.2.3.7) may also indicate atti- and if not, is it society the fishermen's behavior? Compati­ tudes towards fishing. For example, if fish- tion or bility covers a diverse set of issues ranging ing is a low status occupation, psy­ not be very from physical, social, economic, and non-fishermen will probably environment. fishermen. chological concerns to the interested in becoming The key is deterrmining what the fishermen and incom­ of Fishermen perceive as the compatibilities 4.2.5.2 Attitudes Areas where data indicate that Fishing and Alternative patibilities. T.owards potential problems might develop can be Occupations emphasized as critical points in develop­ ment projects. If development or management policies are expected to result in displacing fishermen, 4.2.5.6 Perceived Relative Advantage then it is essential to determine attitudes of Innovation towards potential alternative occupations in comparison with fishing. Data required here would indicate whether or not fishermen feel that they would be 4.2.5.3 Attitudes Towards Income better off in terms of such concerns as their Changes financial position, work schedule, nutrition and/or status as a result of adopting a cer­ Sometimes proposed changes in the fishery tain innovation. The variability of these may have a negative effect on incomes. Peo- data across fishing types, geographical regions, and in relationship to individual incones,pie usually but react it is necessary negatively to examineto decreasing their attributes such as fishing status, a.ge, and of such changes. education is important. perceptions of the impact of edu­ This will facilitate the development cational programs to implement the changes as well as the development of pro- 4.2.5.7 Perceived Trialability of grams to mitigate the perceived impacts. Innovation Once the magnitude of potential income changes have been determined, a sample It is important to determine whether or not survey can be used to assess the impact such the fishermen feel that they could actually

57 try out the innovation if they wanted to and Most of the data identified for collection is why they do or do not believe so. The varia- relevant to analyses of other portions of the bility of responses to this issue should be system and has multiple uses for a variety examined as a function of different fishing of analyses particular to the harvest sector. types and geographical regions, and in rela- It should be clear that this guide is not tionship to individual attributes such as intended to define the entire scope of fishery age, education and fishing status. department's data collection responsibili­ ties. Furthermore, to differing degrees, the 4.2.5.8 Perceived Observability of biological, sociocultural and economic data Innovation identified will be found to be necessary but not sufficient to cnnstitute the department's It is important to determine whether or not contribution to more broadly defined pro­ he fishermen feel that they have the oppor- grams, those which require the coordina­ tunity to adequtely assessthe resultsofthe tion of several departments and/or innovation. Some will accept verbal descrip- ministries. Among these are 1) a thorough tions; others need to witness actual demon- examination of fishermen's welfare and of strations. The variability in responses ways to improve it (see Smith, 1979), and 2) across fishing types and geographical the identification, design and evaluation of regions, and relative to individual variables investment projects (see Campleman, 1976). such as age, education and fishing status should be evaluated. The limits to the data are not so apparent in the first section of this chapter because the 4.2.5.9 Individual Attributes- Asso- fish population is "totally involved" in the ciated with Innovativeness interaction between man and resource. The Perhaps the most important data are those second section, as we have seen, concerns of fishermen; how demonstrate which individual attri- itself with the population which fishing process, its organ­ butes are associated with the willingness to culture affects the how cultural factors might adopt change. Important variables include ization, etc., and education, degree of liter- condition the acceptability of innovations. years of formal to issues to mass media, cosmopolitan- These factors are clearly important acy, exposure welfare and to the ness, social status and degree of social affecting fishermen's attributes design of investment projects. The limits on mobility. Important personality sec­ toward change and risk data will be most clear in the present include attitudes we seek to describe (see 4.2.4.2), gratification orientations, tion. The "population" achievement motivation and level ofaspira- with economic data is the "population" of aonsThieen tivtipontandlev of ai- fishing activities, its state (the private and tions. The relativeimportance of these vari- social benefits and costs of its operation) ables should be determined for the various and its potential response to innovation subgroups in the population. (much of the same data regrouped to lend itself to analyses of cost, supply, etc.). 4.3 ECONOMIC DATA This orientation precludes a discussion of 4.3.1 Orientation the problems of collecting a set of data which is, by its nature, intimately tied tothe It is worthwhile to reemphasize here that operation and response of the harvest sec­ the scope of this guide is limited in two tor, i.e. consumer or retail demand data. ways: first, it encompasses only a portion of Data on the primary sales of fish or other the entire fishery system - the harvest sec- marine products is identified for collection tor. The importance of that sector was dis- below, hence the link between the harvest cussed in the context of the fishery and the sector and the rest of the system will be economy, of a decision process to "develop" made. However, to be able to appreciate the the fishery's role in protein supply and in performance and the benefits society the context of that sector's relationship to derives from the performance of the entire the entire fish delivery system. Secondly, system, data on final demand determinants within this sector the scope is limited should be collected. further in that the data identified for collec­ tion and the information developed from The responsibility for collecting much of them are primarily concerned with the these data, (for example, that on prices of interaction of man and resource: how the substitutes for fish, on consumer incomes, harvest sector operates in biological, eco- the food or general price index) usually lies nomic and sociocultural terms and how it with other departments. However, resp ansi­ might respond to changes. bility for collecting the important time ser­

58 able to ies of retail fish prices and quantities curve and the total cost line. We were lie with the fisheries depart- identify the economy-wide (social) optimum prcbably does pri­ ment. Final demand data are especially on the basis of fishing industry market In other words, important, for example, if by-catch from an ces for outputs and inputs. .industrial fleet competes for the first time the industry determined prices were economy-wide the artisanal catch at the retail level, assumed to be the same as with for some goods Some approaches to the problems asso- (social) prices. In reality, data for demand anal- used in the fishing process, market prices do ciated with collecting cost. But for are contained in Sutinen and Polinac reflect their social opportunity ysis services, notably those (1980). the most important of capital and labor, the market prices are either lower or higher than their 4.3.2 "Economic" Data frequently cost. and social opportunity the harvest sector Given this focus on data shall fishing activity, what economic comes about we review, briefly, some of In fact the equilibrium which be collected? If or which is estab­ of what we have called "economic" in an open access fishery the uses of biological or bioeco­ data, we will see that we can collect a set of lished as a result and can be ana- nomic regulations, is arrived at because data that has multiple uses -to the viewpoints, fishermen respond to market prices lyzed from.different prices they encounter day after day. Taxes, etc. The figures used in the last section of Chap- subsidies, import duties, market power, III show a simplified bioeconomic rela- are reflected in these market prices. A cen­ ter may tionship between the small-scale fishing tral planning unit in the government fleet and the resource. We discussed these see a considerably different total cost curve figures in terms of the relative merits of var- (or even total revenue curve) based on true ious long-run equilibria for the fishermen as (social) economic costs. As will be seen department a group, consumers, society, and the below, however, the fisheries resource. The standard empirical economic can contributeito the calculation of at least cost to analyses of supply, cost, production and one social cost, i.e. the true resource demand data, i.e. data on transactions and society of the operation of the fishery. productive proce';ses, yield the particular information for a given fishery. Another set of analyses, called project or analyses, examine various This information, which is based on investment evaluating the costs and bene­ assumptions of individual consumer and methods for the fits of private and public investments, i.e. producer behavior, is used to determine to and to predict the for evaluating whether it is worthwhile state of the fishery Economic project the fishery to changes, i.e. where move to new equilibria. response of the benefits and costs new equilibria will be. analyses evaluate the from the point ofview of the entire economy. functioning economy Financial project analyses are concerned In the "perfectly" of the return there are no distortions in the optimal allo- with the return and the timing of the nation's resources, including to the investors. Obviously, any "worth­ cation so from the capital and labor, to one use or another. while" project must be judged a whole as well as Some further assumptions are that all par- point of view of society as ticipants are aware of the alternate oppor- from that of the participants. tunities for their capital and labor services, that there is no monopoly (or monopsony) In economic project analyses frequent use is power being exercised in the markets, and made of prices other than those we expe­ that transfer of these services from one use rience in the marketplace - particularly the to another is costless. As a result, the rice ital andtlac e iew te market prices for goods and services in their prices of capital and labor. The view taken current use are exactly equal to their oppor- in this type of analysis is that of a "general tunity cost in any other use in the entire manager" of the economy who is, theoreti­ economy. Because these assumptions were cally, knowledgable about all of the alter­ in effect when we discussed the figures in nate uses of capital and labor in the entire is aware of their true social or the last section of Chapter III, we were able economy; he cost. :' Financial pro­ point of maximum economic national opportunity to identify the among other things, based on the total revenue ject analyses examine, yieldyie__d__MEY)__ased__n__thetot (MEY) __revenue whether a project can be carried out by the

3 This per unit social opportunity cost is often called a "shadow price."

59 participants. They use market prices exclu- as MEY and MSY as "better" in a number of sively in determining the profitability to the aspects when compared to an over­ participants of investing in, for example, a exploited, open access equilibrium, we know new boat or ice house.4 Notice that both that from society's viewpoint it is not types of analyses examine the net benefits alv, ays beneficial to undertake theae moves. of already identified projects - often an Project analysis provides the tools to evalu­ array of projects - which address an identi- ate the economic and financial costs and fled problem. Furthermore they attempt to benefits of these proposed changes. It is not evaluate the flows of benefits and costs of limited to examining the effects of the predicted outcomes of the projects. The investments. accuracy of these predictions is dependent on a thorough understanding of the biosoci- A third use of the economic data is in inter­ of the fishery and its firm or firm type comparisons of private oeconomic operation 7 response to proposed changes. profitability (returns). As noted in Chapter III, the decision to participate in the fishery, In the context of small-scale fisheries two to enter or to leave it, depends upon the particular aspects of project evaluation return each fishermen thinks he can make need to be mentioned.'Both concern what with his capital and labor in its next best are called secondary effects - the benefits employment. He looks at his opportunity and costs which occur outside a project's cost of fishing. What are the quantities boundaries but which are brought about as whose sum must equal or be larger than the a result of the project. The first, the employ- opportunity cost of fishing? Put another ment impact (either positive or negative), is way, how does (or can) the individual fisher­ considered by many economists to be prop- man decide to continue in this employment erly included in economic (but not financial) of his labor, his fishing and organizational project analyses when a particular sector skill and his invested money? There are has an unemployment problem. The second, many accounting measures used to portray which is strictly termed a technological the statue )f a business. The individual externality, is known to us as the project's fisherman is concerned with the annual effect on the resource and the resulting return to his labor, management .nd invest­ effect on catches in the fishery." In this case ment and with his cash flow situation, both financial and economic analyses among other measures. Whether a fisher­ should consider the long run possibility of man, in fact, explicitly does such calcula­ diminished or increased catches as a result tions or not is beside the point. His decision of increased or decreased pressure on the to participate in the fishery is based in part stocks. Economic analysis will take the on the former measure. His ability to meet broader view of the impact of investments expenses on a continual basis is measured on society including the production impact by the latter., on participants and non-participants alike. Financial analysis will consider production The prices used by an individual fisherman changes only in as much as they affect in these accounting exercises are, with a few participants. specific exceptions, market prices. The exceptions in this case are not economy­ In Chapter III we discussed some of the wide opportunity prices, but private, tem­ adjustments that would take place in mov- porary prices. For example, a particular ing from one equilibrium point to the next. fisherman may have borrowed from family While we will always consider points such for some equipment at a very favorable

A widely used reference which examines these two types of analyses is the World Bank publication by Gittinger (1972).

' See Campleman (1976) for a discussion of these analyses applied to fishery investment projects.

We can easily choose some time frame within which the benefits to an investment in the fishery - such as one which increases yields - will exceed the costs. This temporary net benefit or induatry-wide profit is the legitimate objective of many investment projects. To be properly evaluated, however, it must be weighed against the consequences of the eventual long-run equilibrium situation.

7Weuse the word firm here very broadly to encompass the organized enterprise of cast net fishermen to that of much more capital intensive trawling operations.

"The repayment of principal on a loan is not part of the calculation ofreturn but is a significant part of the cash outlay. Meeting a monthly loan repayment provides a great incentive to fishing. In cases when fishing is seasonally dangerous because of weather conditions, this incentive may increase the number of fishing accidents.

60 interest rate or, more commonly, fishermen financial and societal terms. For biological may use family labor paying them less than reasons and because of legal and business market rates.! The prices used in comparing conventions, the year (any twelve month two or more firms are market prices so that period) is the longest and generally most the advantages of having larger families, useful period of time over which we aggre­ for example, are discounted. The prices used gate data in order to make comparisons or in any analysis then, depend on the context recognize trends. Many of the figures in or the scope of that analysis. Chapters I and III show the long run equili­ brium relationship among quantitative eco­ The answer to the question about which eco- nomic, physical and biological variables. nomic data to collect, then, is that we seek Given the proper data set, these figures can data which will satisfy all of these uses be estimated using annual total quantities given the limitations discussed above. We of these variables. As one cannot possibly firms would hope to collect all of the data generated by useseek (ifthe they data kept that detailed individua records) in calcu- each individual firm and all its fishing irmn alis ficom­ return to labor manage- ac ivi lating their annual nor rely on the existence of com­ invested money and in evaluating activity ment and records kept by those who fish, we will cash flow situation; market price data plete their use various sampling schemes to estimate when private and market prices diverge; of the we need in weighted sample means or averages data on certain other quantities (kilograms or money the social or true economic many quantitative order to calculate 2 variables which describe this of this private activity (wasted fish, fcr amountsI) cost to be used will be and data on the physical nature activity.' The weights be ofhexmeffortexample); exndd b the firs.al 10tchosen and the sampling strategy will of the effort expended by the firms.'0° " planned on the basis of qualitative varia­ The data collected will provide the basis on bles such as firm types, species, lcr.ations, which the fishery department can 1)under- etc. circumstances of fish­ stand the financial entire fleet occurs of particular types, 2) monitor The fishing activity of an ing firms time, i.e. over the course of circumstances from year to year, 3) as a flow through these is possible to define a unit of this make comparisons among firm types, and the year. It 4) identify projects or regulations to activity in many ways. For reasons which improve these circumstances. Furthermore, will be clearer as we go on, we define a unit input for the eco- of fishing activity as a single day's fishing these data will provide the smallest analyses necessary by an individualfirm. This is the nomic and bioeconomic used." A trip or state (and potential) of the meaningful unit that can be to determine the set of units of fishing and to predict its response to a landing represents a harvest sector variables whose annual or regulations. activity. For those projects totals are tied to the amount offishing activ­ data collec­ 4.3.3 Harvest Sector Activity ity that takes place, the better tion scheme is the one which, for the same One goal of data collection is to account for cost, properly acc'nnts for the total amount the total amount of fishing activity carried of fishing activity. This implies the ability out by the entire fleet (or by all participants to monitor the distribution of the units of in the fishery) and what this means in fishing activity over time and space.

in rates is called a 9Interest rates on loans for fishing are generally higher than those for "safer" ventures. Thedifference risk premium. and not "prices." We will be 10Note that we said "quantities" needed in order to calculate the social or economic return, prices. But regardless of which prices are used (i.e. which level of analysis we are doing), the gathering private and market wide come from data on the private activities of firms. The decision to use shadow prices (economy quantity values must with planning ­ opportunity prices) is one taken with some consultation with those ministries or departments concerned those which calculate these prices. (in studying production and IICatch and effort data provide the bioeconomic link. It is required by both the economist cost) and the biologist (one of several tools for population studies).

12 We will see below and in Chapter V that if certain data are available we may be able to estimate total annual quantities can have in estimates based of certain variables - particularly those related to catch - with greater confidence than we on sample means.

"3The unit of activity, a day's fishing by a single firm, is the "unit of analysis." This is distinguished from the "sampling unit" which is, ideally, a landing.

61 The proper sampling of activities on this name of the boat owner if different from basis will result in better estimates of the that of the respondent; the number of per­ sample means of, for example, variable sons fishing; the area(s), or zone(s) fished; costs and i evenue and in better estimates of and the trip length.'' The data recorder the weight and species composition of the should be as observant as possible. In an catch. The bioeconom;c compatability of interview setting, for example, it should not biological and economic data will also be be necessary for him to have to question the enhanced if one can calculate catch, varia- fisherman about every identifier. The more ble cost and revenue per-unit activity, identifiers associated with the data, the Furthermore the opportunities for simul- more possibilities there are for making dif­ taneous and/or complementary biological ferent statistical comparisons, for aiialyz­ and economic data collection efforts will ing different aggregations of the data and for supplementing data collected by increase, others. I 1) identify all the In what follows we will variables of interest, 2) define what consti- The respondent may not be willing to iden­ tutes a "good" observation on these varia- tify himself, the owner, the home port, etc. bles as if the situation at hand would permit This reluctance is often related to the exis­ them to be measured in their "best," detail- tence of or threat of the imposition of taxes, revealing setting and, 3) discuss the eco- registration fees, etc.I" In general it is more nomic context of the variables. If the important to get as many data as possible in fisheries situation faced will not permit this situation and to be well received on sub­ observing these random variables in detail, sequent visits than to sacrifice future coop­ then the remainder of this chapter will at eration for a few additional identifiers. least give some idea of what is being missed and how these missed quantities may bias In succeeding sections a narration is used to sample means. Chapter V will examine how identify and define observations of the vari­ to relate this detailed data to the rest of the ables of interest so that the material flows units of fishing activity which have not logically. This narration follows fairly been observed. If one is unable to monitor closely the data collection forms included in the distribution of fishing activity over time the appendices. Slight variations of these and location, then Chapter V will provide forms were used in Central America. some idea of the biases which may be intro­ duced in using sample means (or other mea­ sures) to calculate the desired total annual The revenue we seek to measure is the value amounts. received for the trip. Most of this is directly 4.3.4 Variable Definitions related to the catch and the prices received per pound or kilogram of fish.' 7 Some of it 4.3.4.1 Identifiers can come from sources other than from the sale of catch. If one or more of these other Depending upon th, situation in which the sources (for example, transporting people or data are being colle ted, one should attempt cargo) is significant, regular and wide­ to record or have i corded as many of the spread, some attempt should be made to following identifiers as possible: date, time estimate it. This fishing-related income is and place of collection; the name of the per- more important to questions of fishermen's son responding; the "type" of boat; type of welfare and to investment project consider­ gear(s) used, the boat name and/or registra- ations than to the bioeconomic operation of tion number; the home port of the boat; the the fishery. Its existence may also allow

14By "type" we mean one of the subdivisions of the fleet that the fisheries department hus decided to use. For example, there may he length or configuration types further distinguished by gear type in use.

'5See the appendices for the various identifiers iised on the different sample forms.

Any noticable increase in the data collection activities of the fisheries department is subject to being viewed by fishermen with some suspicion. Some department communication about its activities by leaflet, radio or other means prior to a data collection effort may allay this suspicion.

17 In some fisheries what appear to be non-weight units are used in selling fish, for example "strings" ofgilled fish or baskets or canoe-fulls are sometimes used. There is, in all probability, u weight range within which these other measures fall. It may vary by species and be subject to "inflation" over time, hut with some effort the average weight can be estimated.

62 in the fishery than value of that which is potentially under­ more boats to remain data be supported exclusively counted should be determined. In those would ordinarily can be from catches. From the col- collection settings where questions by the revenue of fish withheld of background information one can asked about the disposal lection be asked. It will often gain some idea of its existence, frequency, from sale, they should only to observe and estimate seasonality and significance.', be possible quantities not sold but valued. It will require the To be more specific, the total revenue some experience to be able to estimate received from selling or using the catch is quantity of fish being so distributed. With of the actual or potential prices experience one should be able to determine the sum ver­ for each qualitative subdivision where and why these various under-or o received of these (species or class) of the catch multiplied by counting possibilities occur. Many weight of that subdivision. For example, potentially under-counted quantities of fish, the above, are total revenue =(price of species 1 (or class 1) especially those in "e" and "f" per pound (or kilogram) X weight of species used to pay for variable costs (see 4.3.4.4, 5, apply estimates 1 (or class 1), plus price/kg. sp. 2 X kg. sp. 2, and 6). In order to be able to In order to get an exact measure of of these quantities to similar fishing activi­ plus...). detailed the revenue (from fishing) for a unit of activ- ties for which there is much less necessary to have an observation on information, it is important to recordsepar­ ity it is var­ species' (or relevant subdivisions') pri- ately 1) the quantities and value of the the 2)the and on the amount actually and/or ious amounts withheld from sale and ces which is sold potentially (see below) sold. quantity and value of the fish in the "normal" transaction, i o.the quan­ There are several possibilities for under- tity and value (by whatever breakdown is counting or over-counting quantities of available) purchased by the middleman or valued fish in trying to get a good measure primary buyer. The pric.s of the fish actu­ of revenue. Over-counting results from ally sold at the landing site are reasonable including the catch of other boats, e.g. pur- estimates of the value per unit weight of the chases at sea or simply the transporting of fish withheld from sale. other vessels' catches for sale. Under­ counting results from: 1) ignoring the sales An estimate of the resource cost of this of some fish prior to the sale at the point of value-producing quantity of fish is the observation, e.g. sales at sea or sales at quantity of all fish killed in order tc get that other landing sites; 2) ignoring quantities of value, multiplied by the current price ofthat fish not sold but valued, i.e. that rejected by wasted fish. If one consider's the dynamic (or simply withheld from) buyers but valued case the cost is higher. This resource cost is by the owner or captain and used a) to pay part of the cost to society of iadividual fish­ the crew, b) to sell at some other site, c) to be ing activity. It does not enter into the pri­ consumed at home, d) to be processed (dried, vate calculations of a fishing firm, yet it is smoked) and sold later, e) to be paid to those an important quantity. The difference who bring the fish to final sale such as between the weight of fish which is killed unloaders, sorters, cleaners, f) to be used to and the lesser amount which generates pay for goods or services used during the value (in money or use terms) results from 1) trip, or g) simply given away. The objective discards at sea, 2) traditional forms of evis­ should be to arrive at a figure (a monetary ceration and other preparation, and 3) spoil­ atirount) which represents the value of all of age (fish landed but discarded). The the fish which is, in fact, valued (all that is quantity of discards at sea is best estimated not discarded) an,1 which is the result of by sea sampling. It is the less or non-valued that particular firia's fishing trip. species which are discarded for reasons other than spoilage.' 9 The form and extent Those quantities which will lead to over- of preparation, including evisceration, is counting should be eliminated, and the usually tied to the appearance (look, smell,

for example, seasonal farming, i.e. from sources not related to the ImWe distinguish this income from that earned by, the are of great significance in examining welfare, entry and exit flows in fishing activity. These latter income sources aspect is opportunity cost of fishing) and the seasonality of fishing effort. This last fishery (they are measures of the of it. Questions of the fishery, and sampling methods we discuss below will take account important to the operation Adiscussion of sampling alternate employment can be included in the appendix on background information. concerning of this guide. procedures to measure this "other" income is, as noted above, beyond the scope fishery can be justified. The 19 unlikely that the cost of sea sampling to estimate discards in the small scale It is very valued species such as shrimp. may be true of sampling discards from an industrial fishery exploiting highly opposite in the last few tows. Shrimp fleets frequently discard relatively low valued finfish caught all but

63 gut contents, dangerous parts, etc.) of nomic meaning in that they are based on groups of species and is reasonably consist- price differentials which reflect how consu­ ent in a given fishery. Biologists work with mers value various types of fish. In order to either whole weight or landed (prepared) maximize the multiple uses of catch data, weights. Conversion factors can he estab- the divisions or subdivisions of catch lished relatively easily at any point before should have biological as well as economic or during the data collection effort. How- meaning. The most commonly used di­ ever, it is essestial to maintain consistency visions of catch which have biological in recording and converting data. meaning (in the sense of relating catch to the population(s) of fish or to a measure of Since fish that is spoiled had commercial potential catches through some population value, the quantity of this fish which is dis- dynamics model) are individual species, carded at sea is probably much less than the cohorts of hidividual species, and, to a lesser quantity which is landed in hopes of being extent, all catch (total biomass yield). There sold or used. The quantity of fish landed but is rarely a one-to-one relationship between rejected for sale or other uses because of the biological distinctions of catch and the spoilage can be very significant in small economic distinctions (e.g. commercial scale fisheries. This fish, part of the classes). resource cost to society, represents foregone revenues to the individual firm. As we have To make this commercially disaggregated said, it is not treated as a cost to the individ. data useful for biological and bioeconomic ual firm. However, this quantity is probably analyses, one can either coordinate eco­ the easiest to observe and measure of all the nomic and biological data collection or quantities that make up the difference apply a recent or soon to follow species between what is killed and what is sold or breakdown (see Chapter V). The accuracy of used. Given this, and the importance of this this latter procedure is directly related to data in suggesting improvements in the how close in time, locatioi, boat size, gear fishery, it is clearly a quantity which should type, etc., the breakdown is to the other be recorded. catches it is being applied to.2' Economic data can be extracted from biologically So far we have discussed the problems of aggregated catch data - such as that which getting accurate observations, i.e., the prob. would result from either a catch and effort lem of over.- ounting or under-counting sampling scheme or one designed to deter­ quantities of fish - principally those quan- mine, for example, length frequencies - if tities of all fish that make up the difference the relevant identifiers accompany the bio­ between what was killed and what was sold logical data. In this case one can determine to the primary buyer. We will also face the the probable prices that prevailed at the problem that the things we want to measure time the biological data was collected and will not be available to he measured in the estimate the costs of 'he trip. preferred subdivisions. The divisions by which fish is sold to primary buyers are Determining the species composition of the based on a commercial classification of the catch is difficult, costly, time consuming catch. Depending on the fishery in question and inconvenient to the buyer and seller of and thenatureof thecatcl,, wt will see sales: the fish. Two hundred or more species may 1) of all fish at the same pe,-unit price. 2)of a often be represented in the catch. If the com­ few commercial classes of fish (two or more position is determined too infrequently, species per class), or 3) of a few important strong biases may be introduced in the esti. species each priced differently and separa, mates of annual totals of species caught. A ted from the rest of the "other" fish. The reasonable compromise is to survey the methods used to distinguish fish in the catches frequently but to sample only a per­ delivery system are usually a matter of tra- centage of each catch. One can begin by dition and are fairly stable though sampling the most highly valued classes (or complex.'2 whatever commercial clasbification is in use) and working through the less valuable All of these commereial divisions have eco- (in terms of price per pound) classes until

P, A much greater degvee of differentiation of fish products is evident in the retail market than in any other part of the system. This often compaic, tes the collection of consumer demand data.

V,Another composite variable we will encounter is nominal effort. We will see later that there are steps we ean take in designing our sampling scheme which will automatically decompose this measure into somewhat more 'conor"tcallyand biologically meaningful components.

64 the annual­ the cla es of fish which constitute about context suitable for calculating 80% of the vah.. ..f.the catch have been /monthly cost to the firm. of sampledl. Because of the possibility data of species in the Two of the more important uses for cost changes in the impartance determining the thorough breakdown should mentioned above are in catch, a more to a firm or a firm type periodically (see (5.2.2.1 (b)). returns from fishing be done and in understanding the demand for process. If each It's obvious that each posi-trip interview inputs used in the fishing with a sampling of the firm consisted of a single fisherman and his will not be coupled of economic data collector, while boat and gear, then the determination catch. The Many fish­ unable to distinguish similar spe- returns would be fairly simple. perhaps however, and, as cies, should be able, or be trained to be able ing firms use crew services classes of the a result, the determinailo of returns to a to recognize the commercial Tli quantities of fish withheld from day's fishing become complicated. different may or may not include a sale. hired crew, which captain (not the owner) frequently share in of the trip just as they share in the The quantitative o', -. vations (weights and the costs 2 prices) and qualitative observations (com- revenues. mercial classes and species) which makt ') and catch are the most If, in a given fishery, the crew traditionally the data on revenue con­ observations to be collected. pay all ice costs, one cannot correctly changeable owner's per­ be different for each individual clude that the firm (from the They will ice in its fishing with each unit of its activity - day to spective) does not consume firm is no demand for ice day, trip to trip. They are also among the process nor that there in the in the fishery. For consistency, and to avoid most difficult to collect, especially the scheme. The variety of either missing or double-counting costs, initial stages of the from for dispensing fish, the speed recording of costo should be separated arrangements these costs. with which it is done, the need to make very the recording of who bears quick estimates of the weight and commer­ fish and many more factors cial class of this Costs - General generate some early frustration. This is 4.3.4.4 Variable will costs which increase with fishing activ­ understandable.detailed data is an The acquired effective skill. collection It is a skill of The fishery to fishery, which requires, at a minimum, a willing- ity, while they vary from aithfisminm ainclude the costs of ice, fuel, oil, :ood, bait, whssoicrereact list can be long. ness to interact with fishermen. labor, unloading, etc. This costs are paid in fish rather Costs in General Many of these 4.3.4.3 than in money. Nonetheless hey are real work is A given firm's net return from a fishing trip expenses. Some background received less all of the required before observations on these varia­ equals the revenues three costs, i.e. those incurred in fishing ble costs can be collected. At least private The first prob­ and those incurred in being able to fish. The problems must be addressed. distinction of costs is between lem is to know what to look for or what traditional which should be variable and fixed costs. Fixed costs are queptions to ask. Items or not the firm fishes. Var- counted as costs will vary by boat/gear incurred whether It wil be costs are those which increase as the combinations and also by location. iable given boat/gear amount of fishing per time period increases, fairly easy to see that a combination uses ice or bait or fuel and oil. Our approach to identifying and describing The more difficult part is to determine what costs rests on the traditional economic dis- costs a firm incu.s in landing the catch and tinction modified somewhat because we in bringing the primary sale to a close. For assume that no records are kept by fisher- example, are there unloading fees or scale men. Consequently we will define variable use fees? costs for a fishing trip and fixed costs for an will describe ice costs, a The second problem lies in determir.ing individual firm. We aways" cost, in the context of an interview where costs end and where "give variable incurred at the end of a fishing trip, and begin. The former are obligations setting goods or services. depreciation costs or interest on loans in a for receipt of necessary

payment. In this section we will count that we counted as revenue to the firm fish which was given to the crew as 22 Recall will underestimate labor costs. this fish as a cost to the firm for crew services. If we fail to do this, we

65 The latter are more properly viewed as choi- catch. Our only recourse is to ask if some ces made by the fisherman in disposing of costs of this trip will be paid later. some of this revenue ecn though these quantities may be dispensed prior to the There may be some difficulty in calculating 2 sale. 3 In some fisheries this disposal of the quantity used of goods which are not income is mixed in with the payment of fish completely consumed during the trip, par­ to meet expenses or fees. Fish used for either ticularly in calculating the amount of fuel purpose need to be observed or asked about and oil used. These are the desired mea­ and then recorded as part of the catch and sures, however. The amount consumed will revenue. That used to pay costs needs to be depend on the boat and motur combination, recorded a second time. the motor's age and condition, the gear used and the traveling time to the fishing site. A third problem is concerned with determin- Thus there are many reasons why fuel and ing whether the size of these different varia- oil consumption for a given boat/gear com­ ble costs is or is not related to the value of bination will differ. The fisherman's esti­ the catch. It is necessary to make this dis- mation of the amount ofeach used should be tinction if we are to keep open the possibility recorded. As with many of the costs exam­ of using several methods for calculating ined, unusual answers can be spotted after annual total amounts of these costs from some experience is gained in collecting post-trip interview data. Some of these other data. One may be able to assist the fisher­ methods may prove to be more accurate man in estimating the per-trip quantity of than the one which generalizes from sample fuel or oil or other inputs not completely means (See 5.4). We w:l! divide the variable consumed during the trip by asking him costs discussed below into those independ- how frequently he purchases the input, in ent of the value of the catch, those solely what quantity, and how many days he dependent on that value, and those whose fishes, on average, between purchases. If magnitude is determined in part by the the prices quoted by the fisherman are sig­ value of the catch. nificantly lower or higher than the inter­ viewer knows to be the "going price" at 4.3.4.5 Variable Costs Independent of other locations, it may be worthwhile to ask Catch from whom the purchase was made and note the response on the back of the ques­ The costs which concern us here are the tionnaire. This will prove helpful should the costs in the first category. They consist of department wish to look into the natui of the cost (quantity multiplied by price) of ice, the relationships between buyers and fuel, oil, bait, food and any other costs sellers of inputs. uncovered which are independent of the catch. The factor which most determines The cost of food is traditionally counted as a which of these costs exist for a given trip cost of fishing. As with all of the coots men­ interview is tbk boat/gear combination in tioned sc far, it should be recorded as if the use. The size of the various costs for any of boat owner (firm) were responsible for it. these combinations is a function of the Adjustments can be mad !ater. For this length of the trip, distance to fishing area, cost, more than any other, an "educated etc.2 4 guess" may be the best estimate. Since the arrangments for purchasing food are so var­ There may also be catch-independent costs ied, it is probably not worth the time to try to which are more closely related to the fact get an accurate measurement. It is easiest to that there was a trip - for example a flat fee settle on a money amount of the value of the for unloading - than to the nature of the food for one person and multiply by the total trip, for example the cost of fuel consumed, number of people who went fishing. One Most of these costs will hav- been paid prior should include in this amount any food to tofthe thep.Fequetwl trip. Frequently, howeesomehowever, costs arsgnralmsbsatilinreaiotoshbrought on board by the crew. If food costs to the for goods or services used during the trip are are generally substantial in relation paid afterwards, even after the sale of the gross receipts from the trip then more effort

2:Strictly speaking this "distribution" of catch is not the same thing as the economic concept of distribution. Economic distribution is concerned with production possibilities based on initial endowments and with the share of output the productive factors receive for their services (See Samuelson, 1973 or any edition).

24 We should be sure that the identifier (or nominal effort measure) "number of days of fishing" is associated with each observation on varifble costs. In the discussion of effort data we will see that more specific data on the effort expended will allow us to approximate the cost of "effective" effort.

66 fish preparation, landing should be invested in determining the true use, brokerage, of these costs is present for cost. taxes, etc. Which any landing is probably more related to the More Two additional costb which are properly landing site than to any other factor. in this section are random costs: frequently than with other costs, these are included sec­ the costs due to losses of, or damage to paid with some of the catch. Recall the gear, hull, sails, motor, etc. One tion above where potentially under-counted equipment, The value can ask if there have been any losses or quantities of fish were discussed. during this trip. What spe- of these quantities of fish was then classi­ damage incurred now be lost or damaged, and the esti- fled as revenue. These values must cifically was costs. The replacement or repair costs (parts accounted for in the payment of mated these costs is to and labor) should be recorded. The labor easiest way to determine as the actual ask the owner/captain about them. Alter­ costs should be recorded who and again, but not as regu- natively one can question all those amount paid conven­ the market would value those servi- were paid. This approach is more larly, as Frequently it ces. All of these costs should be recorded as ient for the captain/owner. would bear the entire amount. will be necessary to settle for merely observ­ if the owner of fish and For an alternate treatment of damage costs ing the often rapid distribution dispensed. Some time se,! section 4.3.4.10. noting the quantities must then be spent at the site to discuss the There may well be other catch independent distribution (distinguishing payment of costs to be considered which have not been costs from the distribution of revenue) with mentioned here. The background informa- whomever is available to discuss it. tion exercise and an increasing familiarity fishing operations of the fleet The reason for asking about which measure with the of the should make one aware of their existence of catch (for example, total value they can be handled. Remember catch, value of that sold to the middleman, and how as we the goal is to identify and get good etc.) these costs are based upon is, that the option of detailed measures of these costs and have said before, to keep open The revenue (value) and the varia- having alternative methods for calculating revenues. costs. This ble costs identified above and those to be total annual amounts of these below are defined for interviews holds for the fourth priority as well. These identified until a repre­ at the end of fishing trips. Some ofthe data questions should be posed only estimated either by visual sentative sample of responses from the dif­ will have to be is (e.g. weight and commercial ferent landing sites is acquired. It observation which these of fish), through post-trip conversa- probable that the basis upon class rule for deter­ with the fisherman (e.g. quantity of costs are determined and the tions so much as tu be fuel and oil consumed) or by more extensive mining them will vary when costs and useless to us. However, one should make the post-trip interviews that hidden by complex relations effort to get a sample before making revenues are these costs is creditors. As much of the calculating determination. If the basis for with landing sites or if work as possible should be done before leav- fairly consistant across site. Simple, inexpensive the number of sites is small anditis possible ing the interview in the post four-function calculators make these calcu- get a consistantly good measure and provide and incen- trip interview setting of the quantity of fish lations less tedious the being done at the site. Standard on which these costs are based, then tive to their measures units such as liters, kilograms, etc. should chances for more accurate annual be agreed upon before the data collection are improved (see 5.4). so that decimal amounts are not mis­ starts Costs understood in later coding work. 4.3.4.7 Labor of labor or crew services for a given Costs Which Vary with Catch The cost 4.3.4.6 trip is somewhat more difficult to determine Alone than other costs. It involves getting good (usu­ data on these costs measures of some of these other costs The priorities in getting and of the what these costs are, ally the non-catch related costs) should be to 1)discover c(st is calculated their size, 3) find out what mea- value of the catch. This 2) determine of the owner(s) of the of cat 2h they are related to and 4) deter- from the viewpoint sure gear. It is the sum of the value mine the relationship. boat and (money and fish) paid for all crew services regardless costs which depend only on catch may (including those of the captain, The the boat owner) less include the costs of unloading, sorting, scale of whether or not he is

67 whatever costs they have pnid or will pay in two for the "boat" (the owner(s) ). The connection with the fishing trip. 25 shares might be based on the total value of the catch, some of which might be sold at The systems, rules or arrangements used to sea before the landing, some of which is determine how much the different partici- withheld from sale for various reasons. The pants in the fishing trip will receive are two crew shares might be "net of" ice, food, called by various names including "share" bait and one-hplf of the fuel costs. In other systems or "lay" systems. Share systems words, these costs are deducted from the are usually established by tradition and two-fifths of the total value of the catch modified as newer technologies are adopted intended for the ordinary crew. The captain (as the balance of labor and capital in the may be responsible for the remainder of the operation changes) or as the risk involved in fuel costs and the "boat" for the rest of the the fishing venture changes. These systems costs, e.g. unloading. While this system is of usually vary by firm type (boat/gearcombi- moderate complexity it could be nation), by location of the home port of the "explained" to the recorder in less than boat but not by landing site (ifit differs from clear terms. If one wishes to record a des­ the home port). cription of these systems it is helpful to ask The authors encountered approximately thatthe rule the forsystem dividing be unscrambled: all costs involved to ask and for revenues (see different systems for determining then for the rule governing thirty approach will usually crew shares in the Gulf of Nicoya fishery in below). Use of this a more detailed description. Costa Rica. All obligated the crew to pay a result in percentage of different (catch independent) trip costs (a certain cost to the crew) and a Should share system descriptions be used as right to a percentage of the catch's value (an ahouid e sstem dero s fo a uncertain revenue). In this way all who a guide to determining the labor costs for a fished and all who owned capital in the firm particular trip or for reducing interview shared in the risk of the fishing venture; in time for future trips or in devising ways to this case the risk of small catches. Just as estimate labor costs for unobserved trips? fishermen enter and leave the fishery in We think the answer to each question is no. relation to their opportunity cost of fishi~ig, Because these systems are so varied and are so, too, crews of artisanal boats are aware of particular to both boat/gear combinations share arrangements in other ports or on dif- and to the home port of the boat, it is safer to ferent boat/gear combinations in the same measure the labor costs directly. These rules locale. No one arrangement is too inconsis- may be used to indicate that we are not ask­ tent with the rest in terms of the potential ing the proper questions. After conducting a reward to fishermen for their labor and risk. reasonable number of interviews, one can If one were, movement to or away from that see if the amounts received are greater than arrangment would be expected. These des- those calculated from the rules. This might criptions of share systems should be reveal that a greater amount of fish is being recorded. They provide useful data for ana- sold distributed before the sale than is lyses of risk bearing, of economic distribu- being reported by the fishermen. tion to the factors of production, ofthe social organization of the fishing activity and of In order to calculate labor costs for a partic­ potential socioeconomic changes to crew ular trip one must determine 1)whether the and firm arising from technical captain is the owner, 2) the amount of the innovations.16 shared costs paid by all of those in a particu­ lar payment category, and 3) the total A hypothetical share system might have revepue (value) paid to everyone in a partic­ five shares to be distributed; two for the ordi- ular payment category. Ifthe captain is also nary crew (captain excluded) regardless of the owner, his payment of costs and receipt their number, one for the captain (regard- of shares should appear in two categories, in less of whether it is the owner or not), and one as captain and in one as owner. The

25 The share of revenues to be received and the share of costs to be paid are different, usually, for different roles or jobs. Frequently encountered divisions are 1)the boat's shitre (representing the owner), 2) (possibly) the gear or motorowner's share (if these things are not owned by the boat owner), 3) the captain's share, and 4) the crew member's share. Theowner who is also the captain therefore gets a larger return than the owner who is not captain.

26 If such studies are anticipated, these systems need to be investigated in detail in some other setting as they are often complex. The simple, verbal communication of these arrangements at the end of a fishing trip will often lack elements such as the obligation of the crew to provide free labor and perhaps parts for maintenance and repair.

68 owner's costs and revenues can be calcu- trip's costs and their division were recorded. lated after the interview is over as these amounts are "what is left" of the total costs 4.3.4.8 Other Considerations and total revenues after all payments have beenben subtracted.sutlractes2'7 Thefter netn paymentpayments totae- peo- Beforeafwipratpit closing this section onedtvariableemncosts determinedple in the various by paymentsubtracting categories from cantheir be tioned.a few importantThe first, andpoints most need important, to be men­ con­ deterned yosbtacting from t r cerns the use of family labor in the fishery. revenue any costs they have paid. This is properly a study in itself. The major collection, for economic an attempt is made to obtain a verbal implication for data If operation of the harvest sec­ description of the share system while deter- analyses of the it is necessary to be sen- tor, for project evaluation and for interfirm mining labor costs, labor is the ordering of the questions comparisons is that this family sitive to both the service of and those whom we question. In other often underpriced. Frequently avoid undermining the family members, especially that of wives words one should the same of respondent(s) by appearing to and children, is not rewarded at credibility services rendered by "checking up" on his (their)responses by rate as are the same be this may be to the pri­ sking both the captain and the crew the hired laborers. While Verification of amounts vate advantage of a particular fisherman­ sante quections. measure of is important and can be done /owner, it obscures both the true received or efficiency in com­ discreetly. that firm's profitability parison with others and the market value of In one method used in Central America the labor services in the entire fishery. How this interviewer isked each of the ordinary crew problem is dealt with in order to maintain members how much he had been paid prior the accuracy of the data and yet minimize to any deductions for the trip (including the the inconvenience and cost of that accu­ of fish receved) and then how much racy, depends upon how widespread this value to he had paid or wjuld pay in expenses for the usage is. First of all, it is necessary trip (including food costs if he paid for it inquire about family member participation himself). Alterlatively each member of the in both the background information exer­ at the ordinary crew could be asked how much cise, and, initially, in the interviews (net) he had gained or would gain from this end of fishing trips. During an interview, be trip. The captain was then asked about the the amount actually paid should total value of the catch, how much each divi- recorded as zero if there is no remuneration. sion of the crew would receive of this Afterwards, the roles family members have amount (including fish) then the general played in either the fishing or in bringing rule for dividing revenues. Afterwards we the fish to sale should be determined. Iffain­ recited a list of possible costs, beginning ily members are involved only occasionally, be with trip independent costs, and asked the then these questions should continue to total amount of each, whether it was a included in the interview. If there is exten­ shared cost and lastly if we had missed any sive use of family in established roles in, for other costs of the trip. The captain was then example, unloading, thcue is no need to con­ asked the general rule for dividing costs. tinue the questioning. But this use has to be One can see from this that we are not inter- documented so that market values can be any eco­ ested in how much a particular individual attributed to those services prior to paid for, say, ice, but rather in the total cost nomic analysis. the eachof the payment ice and howcategory. much of it was paid by Secondly, we have tried to ininimize time and inconvenience of the interview by which In another method, reflected in the trip omitting questions, the answers to questionnaire in the appendix, the inter- can be deduced from already recorded data. questioned the captain exclusively. For example, the return to the boat owner viewer the The set ofquestions asked was similar to the for a given trip is the difference between above except that when a shared total revenue (value) received less all of the set asked of an was identified, the amount that each variable costs. Likewise the salary cost calculated division of the crew actually paid (or would ordinary crew member can be pay) wtas recorded. The general rule for costs once the number of people ii that category was determined only after the particular and the net amount of the revenues receivd

can 27 In some cases it will be found that an owner is providing services such as housing, food, etc. to crew members. We the annual costs to the owner should uncover this situation in our background survey ofthe fishery. Ifthis sitration exists be determined and treated (sampled) as a fixed (labor) cost (see 5.4).

69 from all people in that category have been ious components that contribute to the recorded. These calculations can be done in physical production process of fishing. the fisheries office. When they should Le Observation3 on these components contrib­ done depends upon how the data are to bz ute to the quantity of data available to deter­ stored. If computer facilities are available mine fishing effort. The greater the which can store and manipulate the data, specificity of the effort data, the closer we then the need to do the desired calculations can come to approximating effective effort as soon as possible after the interview is and its cost. Prior to an interview the reduced. What calculations should be done, recorder can observe the boat type (and of course, depends upon what the data will appi oximate its length and width), the be used for. It is worthwhile to do at leastthe number of people who fished, the kind of obvious addition, subtraction, multiplica- propulsion used (sail, motor, etc), and some tions and division necessary to complete the of the gear used. In the interview one can description of a given fishing trip eg., total determine the total length of the trip, the revenue, total variable costs, labor costs, specific kinds of gear used, the actual fish­ etc.. Regardless of how the data are stored, ing time each gear was employed, the area the ability to generate interesting data from in which each was employed, and the a given interview weeks or months after the motor's power. Our experience has been fact, or to generate data (e.g. sanle means) that most fishermen are willing to talk by aggregating over many interviews, is freely about almost all of these details directly related to the number of consis- except for the exact locations fished. They tently recorded identifiers which accom- have been willing to indicate in which zone pany the data. they fished, however. Thirdly, it should be obvious that compiling After gaining some familiarity with the the suggested data in the detail we have kinds of gear(s) employed, it is possible to specified is not easy. The burden of gather- prepare the questionnaires to include a set ing this detailed data will fall on those plan- of specific questions for each gear type ­ ning the sampling scheme (Chapter V) but questions seeking detail on the fishing more so on those actually collecting the power of the various gears (see the question. data. There will be a tendency to estimate naire in Appendix C). For example, some amounts (or worse yet, guess at amounts) of questions specific to gillnet use concern how catch revenue and costs when they should it was hung, the stretched-mesh size, its be properly measured by interacting with area (length and height), and the number of the fishermen. This tendency will increase meshes between ties on the support line. In as data recorders become more and more general, these questions address those familiar with the data. aspects of the gear and its use the investiga­ tors and the fishermen think contribute to Lastly one should be aware that the amount its effectiveness. Every aspect of fishing of detail specified in defining a good mea- effort contributes to either the economic or sure of revenue and variable costs and the biological significance of the firm's fishing amount of time and energy that we suggest activity. Many contribute to both. We will be spent in getting that detail, reflect the miss some interesting economic data if we authors' experience with the magnitude of fail to record, for example, the motor size the cost and revenue elements. It is very (H.P. or displacement), simply because the likely that the relativesize of these different fisherman uses a stationary gillnet. costs and of the different amounts of fish that constitute the total value of the catch Because the number of days of fishing will will be different for different fisheries. One be difficult to determine for any one firm should keep in mind the multiple purposes type and/or the cntire fleet, the number of which the data will serve, the amount of days fished in the previous 30 days should detail this requires and then allocate availa- be determined. ble data collection resources on the basis of relative magnitude of the quantities Once again, it is helpful to have established the con­ involved, a set of standardized measures to avoid fusion in recording and interpreting data. to establish classes 4.3.4.9 Physical Production and Effort Specifically, it is helpful of boats and gears, fishing zones, methods Data of employing gear, materials used in the Post-trip interviews can be used to collect a gear (e.g. mono- or multifilament line), and reasonably detailed description of the var- standard units of measurement.

70 4.3.4. 10 Maintenance and Repair Costs By including maintenance and repair ques­ tions in post-trip interviews we make them The cost of maintenance and repairs is usu- longer but gain detail and accuracy. If this ally significant for the individual fisher- approach is nacceptable for whatever rea­ man. For a given firm this cost increases as son then one should at least be aware of the the amount of fishing in a period of time biases which are introduced in extrapolat­ increases. Regardless of how much fishing ing from sample data to annual total is done, some expense is always incurred amoun-ts of these costs. The less mechan­ because of the need to combat the effects of ized the fleet, the more likely that mainte­ sun, rain and sea on assets such as the hull, nance costs will be larger relative to repair sails, motor, and gear.2 costs, and the more likely that these costs will be incurred during a short period oftime How can tlS annual total of this cost be - perhaps in a season when fishing is dan­ Should it be separated into two gerous. If this is true then it may be possible componentsdetermined? - regular maintenance, to arrange the fixed cost data collection iniie cotentali on a schedule) and randomly occur- to (implying to minimize potential biases. ring repairs? How shall it (they) be schedule records of this sampled? Since written fol­ costs are lacking, one must depend Let us review the necessary data. The (these) equip­ the memory of the fishermen. If mainte- lowing data items apply to all capital on for one major nance is separated from repair costs we may ment. They should be recorded with a distinction that is not item at a time, i.e. one should ask all the be working and repeat recognized by the fishermen. While there is questions concerning the hull and again for the no general rule, we suggest considering them for the motor or sails as one and treating this one cost as a different gears used. During the post-trip them indicating variable cost. Damages, of course, occur interview one should begin by but we suggest treating them as interest in all maintenance and repair costs randomly by these costs as well. This was done, in since the last trip(including that caused part of for fact, by including damage costs among damage). It is worthwhile to suggest, trip-related, catch-indepen dent costs each major capital item, several likely those One above, repairs which might have been made. needs to record the item or task under dis- Ar3 the costs of keeping the boat in working cussion, (for example, caulking the hull) the order incurred as they arise or are they actual cost of materials, and if the costs (aside from damage costs) put off until a were shared, and what fraction was paid by slack period of fishing activity permits the each of the groups of people who shared in work to be done at a smaller real loss of the cost. One then inquires about the actual fishing time?29 The answer is probably that total cost of labor, how many hours (or pos­ those things which need to be done are done sibly days) were involved in the work, and as they arise and that recaulking, repaint- again ifthe costs (theamount actually paid) ing, etc. is done in relatively slack periods. were shared and how. One should then ask The question then becomes which sampling who provided the labor: was it hired or was procedure - one which samples fishing it done by family members or crew? The trips or one based on the population offirms cost of labor may well be zero. If so, it should (represented by fishermen at home) - will be recorded as zero. However, if labor was capture these costs better. It is suggeeted provided by the crew or family members at that the better procedure is the former no or reduced cost, a measure ofthe market because post-trip interviews will be more value of those labor services should be evenly distributed over time and type than obtained, i.e., the hourly market rate or the sampling of firms for fixed cost informa- what the entire labor cost for a task would tion. Furthermore, a set of consistent ques- have been had someone outside the crew or tions can be posed which will, in effect, family been hired to do the task must be cover the entire year's distribution of main- ascertained. It need not be done for each tenance and repair costs and yet not tax the interview, however. Having recorded what fishermen's memories for detail. the work was. who performed it and the

28 Depreciation costs are accounting usts which can be evaluated at any point in time For this reason they are considered to be fixed costs and are sampled as such.

29 Maintenance and repair costs are amounts paid to keep the boat in condition to fish. We should distinguish these from costs incurred for major renovations or refitting. These latter are more properly considered reinvestments of capital and will be reflected in the new (larger) current value of the assets. The cost ofthese investments is discussed in the section on interest and depreciation costs.

71 amount of time involved, we can indepen- If fishermen keep detailed records, mainte­ dently obtain market rates for the type of nance and repair costs (including damage) work. Boat yards or boat builders are good should be determined once a year. These sources for this information, records, if measured from a proportional sample of the population of boat/gear com­ The market valuation of labor eervices is binations, will yield a good estimate of the used differently if the labor was supplied by annual maintenance and repair (including the crew rather than by family members. damage) costs (see Laxenaire, 1973). In the The market value of the family's contribu- absence of log books or accounting forms, tion should be included as the labor cost of one should try to adjust the timing of the the repairs. If the crew provided the services fixed cost sampling, as much as is reasona­ "free of charge," we assume that this was ble, to capture as much of these costs as taken into consideration earlier when crew possible. This is easiest if these repairs are shares were negotiated. Therefore there is done in an interval of lessened fishing activ­ no charge to the firm for labor in this ity. If this work is spread out over the year, instance as the cost has been essentially then an adjustment for capturing these prepaid. Nor is there any deduction from the costs (in terms of accuracy and not logistics crew's share of the trip revenue. In this rase and sampling cost) is to distribute the fixed the market value of labor services is a ,ece cost sampling scheme in relation to the of datum not used in calculating trip profita- amount of fishing activity that takes place bility. It is used in other analyses as a mea- from month to month or quarter to quarter sure of the expected labor costs for repairs. and to assure that each type of firm is The share of parts costs paid by the crew sampled in each month or in each quarter. If should likewise be deducted from their trip one can assure this proportional representa­ revenues., There are no distortions created tion of boat/gear combination in each in the estimated annual amounts because month or quarter then one can ask that these previously incurred costs are deducted costs be remembered since the last month or from a snlbsequent fishing trip's revenues, quarter respectively. Finally, one should ask how many fishing were lost since the last trip due to Where no records are involved, the bigger days repairs repairs. This loss offishing time is not a cost problem will be in recalling what foregone revenue. As such, it does not were done, what parts purchased, the dura­ but It is enter into the annual costs of fishing. tion of the repairs and fishing time lost. likely that the rule for sharing costs, the were Since relatively significant expenditures on probability that family or crew maintenance and repairs (including dam- involved, the market price for labor services repairs) are important to the fisherman, and even the price of parts will not have age year or they may be mentioned even though they changed much over the course of the prior to the last trip. It is worth- quarter and certainly not the month. On the occurred over­ discussing these repairs but we other hand many small costs will be while cost setting. should record only those costs incurred looked uLing this fixed since the last trip. Otherwise, we will over- We will now appear to contradict ourselves count this already importan. cost. by recommending that both procedures be used simultaneously in the 0irst few (two or When one attempts to get data on mainte- three) years of data collection. Because of nance and repair costs in a setting suited to the many factors affecting the occurrence of collecting fixed cost data, the time reference these costs, it will take perhaps five years to the questions, which are otherwise before arriving at a reasonably accurate exactly the same, changes. For example, the estimate of the annual cost of maintenance time interval in the following phrase is vari- and repairs based on post-trip interviews able. "Let's discuss maintenance and repair (see 5.4). This procedure will pay off in the costs you have had in the last long run, but much can be done to augment months." What number one puts in the data on these costs in the interim. We might blank space, as we have said, depends on add the costs collected in each procedure for the relation between the distribution of the first year, take an average of the results these costs and the distribution of our fixed ofeach procedure the second year and grad­ cost data collection efforts through time. In ually decrease the role played by costs col­ fact, it is e', en more complex if we have a lected in tb2 fixed cost procedure in variety of boat/gear combinations which succeeding years. Alternatively, and vary greatly in the amount of capital each depending on the fishery, we might initially employs, base our estimates of these costs on infor­

72 require no mation gathered from boat building and ing methods. The questions the recorder in mat­ repair enterprises, training on the part of ters of business or eccnomics. about fees 4.3.5 Fixed Costs One should first ask questions of moor­ costs are by convention annual costs. such as the amount and frequency Fixed fees and Some of them may be paid for weekly or ing and protection fees, cooperative intervals but they so on, and then begin the discussion of monthly or at irregular For and compared on an annual investments and capital equipment. are discussed a descrip­ basis. When collecting data on these costs each piece of capital equipment, be asked about the year that tion of the item is needed. This description questions will the hull ends on the date of the interview. The sam- should be as detailed as possible: costs should be carried out should be described by length, width, draft, pling of these mate­ within a given year (any twelve month capacity of the hold and construction Regardless of how the sampling is rial; the motor by brand name, horsepower, period). gear by its distributed throughout the year, the data type of fuel used, etc.; and the to the 12 month period ending with material, dimensions, and so on. Record will refer when it collecting of the last fixed cost data. If information on the age of the item, the age when it one had written records of these costs, for was bought by the fisherman, accounting sheets, one could com- was bought and his estimate of remaining example with questions cost data for the same time period for all working life. Follow this pile historical and present firms. This suggests that a more realistic about the item's total fixed costs value. One needs to determine the price of picture of the fleet's annual he by carrying out all of the the item when it was purchased, the price could be gained ofthe for these costs during a very would sell it for now, and his estimate data collecting identical in char­ short period. Given limited manpower this price of a newreplacement This work should be con- acteristics to the one he now owns. Deter­ would be difficult. major if there is a period in which mine whether he has made a centrated only lengthen­ fishing takes place. The regular- reinvestment in the item such as very little adding new with which post-trip interviews and ing the boat, replacing the deck, ity the approximate landings surveys are conducted should not superstructure, etc., and be interrupted if there is no period of greatly date and cost of these renovations. diminished fishing activity. For major items such as the hull and motor, of the Fixed costs can be divided into two groups: one should obtain estimates outside for services independent of interview of the perceived present value. cash payments should the amount of fishing activity (such as Local merchants or other fishermen recent sales of mooring and protection fees, cooperative be interviewed concerning membership fees, accounting and legal similar items in that location. and those related to the use of expenses) costs can be determined from capital such as interest costs - also a cash Depreciation the data above. These involve calculations payment - and depreciation -an accounting will the annual which require some expertise, and cost. For most of these costs the analysis. from responses to depend on the purpose of amounts are calculated of calculating depre­ weekly, monthly or quar- There are several ways questions about for any capital asset. A private payments. ciation costs terly cost can be calculated based upon the origi­ offixed costs will nal puchase price and some standard depre­ Under this broad category or related variables ciation methods such as the straightline be several other closely such to the fisherman's cash the declining balance method. Assets which are relevant at different situation and important in the consid- as hulls and motors depreciate flow they have different life eration of investment projects. For example, rates in part because questions about indebtedness, the amount expectancies. in the fishery, (value) of capital equipment assets of the individ of boats, motors and gear Regardless of how the and the durability are depreciated, he must face will be included. The proper setting for col- ual fisherman there is market (current value) prices in replacing lecting these data is one in which depends interview and an them. The current value of a boat adequate time for the those which are of privacy for the respond- on several factors including atceptable level the private depreciation interviews can be combined with used in calculating ent. These initial value, present condition aimed at gathering either sociocultu- cost, i.e., its others However, the current market value ral data and/or more detailed data on fish- and age.

73 is also influenced by the general rate of case of interest, the true private cost is very inflation and what economists call the likely greater than the market cost (the value of the marginal product of capital. bank rate). It may be that fishermen do not This last factor is influenced by the state of have access to this regular lending market the fish resource and this, in turn, by the for a variety of reasons. If this is so we supply of boats. Consequently, one should should use our calculation of the cost of bor­ not be surprised to find that the current rowing (which is based on the rates they value of the boat and motor and perhaps of actually pay) for any analyses particular to the gear, is in fact, greater than the original the fishery. cost of the item. A portion of this larger amount may result from improvements in 4.3.6 Industrial Fishery Data the item. To discount this effect, questions on reinvestments have been included. With- The industrialfleet is guided by the same set out going into this further, one can see that of economic forces which affect the arti­ theit is currentworthwhile value toof get the several boat - opinionsaside from on persanal firm fleet. is muchHowever, greater the capitalin the formerinvestment and, the one given by the fisherman himself consequently, the opportunity cost (what of an he says he would sell for now). industrial fishing venture is dominated by the alternate uses of this capital. The same Interestcosts are cash amounts paid for the kinds of biological, physical and economic use of another's money capital in order to data which describe small-scale fishery purchase capital goods such as the boat, operations also describe the operations in motor and gear. As with depreciation costs the industrial fishery. The mi-jor difference, these calculations require some expertise. from the viewpoint of data collection, is that To determine capital costs the following a greater percentage of most kinds of data is should1. be recorded for each capital item: recorded by individual companies. For a whether a loan was involved in purchasing thorough analysis of the industrial fishery, the item; when the loan was made (as pre- the same economic, biological and physical cisely as possible); who, or what type of data need to be collected as are recom­ lender supplied the money e.g., family, mid- mended for the artisanal fishery. dleman, a money lender, a bank, etc.; what the stated hmount of the loan was; how We will concern ourselves here with the much was ac'ually received; the stated interaction of the two fleets. This interac­ interest rate; the duration ofthe loan or how tion can take place in the input market, at many payments are yet to be made; the fre- any place in the output market and in the quency of payments (weekly or monthly); fishing process itself. The impact on the the part of the payment which represents artisanal fleet of participation by the indus­ repayment of principal; that part which is trial fleet in the input market can be both repayment of interest; whether these pro- positive and negative. The kinds and portions are changing over time and how; amounts of inputs available to the artisanal whether the periodic payment is increasing fleet may increase because of the emergence orthe decreasing loan, if any, in size;and itsthe naturecollateral and used value, for thisof an the industrial amount fishery. of infrastructure We can include (ports, in piers, etc.) and newer fishing technologies. The private interest rate quoted by the On the other hat the prices of certain fisherman may or may not correspond to inputs such as fuel, ice and money may rise the rate he is, in fact, paying. The market if the amounts supplied do not keep up with rate quoted by banks, while reflecting a risk the joint demand. premium, is usually standardized in that it is based on the nominal amount of the loan In the output market - from the primary and holds for specific periods of time. The buyer through the retail level - large firms rate quoted by the fisherman may be lower may supply their target species or by-catch than the bank rate and this, in turn, may be to compete directly (same product) with the lower than the rate the fisherman is actu- output of the artisanal fleet. This will yearally paying.includes His the interest (one coststime) fordifference the first depressdemand pricesdoes notto the keep artisanal pace with fishermen the joint if between the nominal amount of the loan supply. An increased supply of substitutes and the amount actually received plus the for the small-scale fishery's output such interest as part of his regular payments. canned tuna or mackerel would also depress fresh fish prices ­ cauding an out­ Up until now, the private costs discussed ward shift rather than a rotation of the have been less than market costs. In the supply curve. On the other hand the indus­

74 trial fishery may open opportunities for pro- nomic analyses. The commercial fleet will, cessing fish from the artisanal fleet, for in all likelihood, keep fairly detailed records creating export markets into which their of locations fished, actual trawl times, etc., supplies can be funneled or for creating an - several of the determinants of effort. The entirely new artisanal enterprise such as quantity and value of fish killed will, in supplying baitfish to the industrial fleet. most instances, be much greater than that landed and this greater than the amount sold. In the catching phase the tw9 fleets may be Sea sampling will be required in order to exploiting the same unit stocks of fish or begin to estimate the amount and composi­ different life stages of the same species (this tion of fish killed and, therefore, not availa­ production interaction works both ways, of ble to society nor the artisanal fleet. course). The social resource cost is the amount and value of fish killed in the opera- The tools of project analysis are useful in tion of the two fleets. There may also be gear evaluating the damage or good caused by conflicts or simply crowding on the fishing fleet interactions. It is unlikely that any grounds. absolute decision can be made in favor of one fleet or the other since society and con­ We should be able to identify, in a general sumers are involved in the benefits and in way, points of cooperation or conflict fairly the costs of the joint operation of the fleets. easily by simple observation and discus- Any analysis must weigh, for example, sions with key informants such as expe- reduced costs to consumers, employment rienced fishermen, input suppliers and impacts on the fishermen, increased or others in the delivery system. A detailed decreased long run catches, etc. Once again, study of the effects of competition in any of economic analysis provides a backdrop the markets requires, ideally, a compatible against which essentially political deci­ time series of data on prices and on the sions can be made. amounts supplied or demanded by both groups. Commercial classifications of fish (not species) are required for an analysis of supply and demand in the output market. Regardless of where the industrial catch or by-catch enter,3 the delivery system, the impact of the increased supply will be reflected in prices paid to the artisanal fisherman at the primary buyer level. If a small scale fishery data collection scheme is in existence, the quantities supplied by the artisanal fleet and the prices will be known. What is lacking then is an estimate of quan­ tities supplied by the industrial fleet. The obvious source of this data is the industry itself or, if a receipt system is in existence, the purchasers oftheir output. Any measure of the quantities supplied by the industrial fleet in a period of time is better than none. In several Central American countries, receipts from purchasers of the highly valued target species of the larger boats included space for a crude breakdown of the incidental catch sold. Likewise, trip reports made available to the fisheries office included amounts of by-catch landed (but not all that was killed). If there is fleet interaction in harvesting it is imperative to attempt to get detailed data on both the quantity and species breakdown of the industrial catch. Otherwise, stock esti­ mates or eotimates of sustainable yield based on artisanal data alone are much less reliable, perhaps even useless, in bioeco­

75 Chapter V

DATA COLLECTION METHODS

5.1 INTRODUCTION observations were biased by the attitudes and vahes of the observers. Ifwritten mate­ The purpose of this chapter is to describe rials are reliable, they can be used to check methods and procedures which can be used the validity of statements made by key to collect the data which were described in informants. Detailed and accurate studies Chapter IV. are not common, but if any exist, they should be located and used as long as their The required data can be obtained using six reliability can be verified. basic techniques: 1)examination of availa­ ble records, publications or other written 5.1.2 Key Informant Interviews interviews, 3) material, 2) key informant in­ receipts, 5) sample This type of data collection technique census surveys, 4) sales with research activities. Follow- volves intensive, detailed interviews surveys and 6) they general description of each of individuals who are selected because ing a brief, information. these techniques and a discussion of inter- can provide extensive, reliable methods, specific data collection For example, a great deal of important viewing fishing activities methods which generate biological, eco- information concerning sociocultural data will be exam- can be obtained from a few, well informed nomic, and and fish mer­ ined separately. local fishermen, shop keepers chants. Additional sociocultural data can 5.1.1 Written Records often be obtained from religious or govern­ ment officials who keep records on births, Where possible, all available written deaths and commercial establishments. As records should be reviewed before any data noted above, key informants can also be are collected. Frequently, written materials useful in identifying sources of written exist, but can be located only after inter- records. Finally, key informants provide viewing key informants (see section 5.3). information which can be used to develop These materials include reports completed questions that should be included in more by outside agencies, studies done by local structured sample surveys and interviews. or students and statistics University staff is an as historical catch and effort data com- Identification of key informants such require several piled by the national or regional fishery offi- important task which may to data collected by previous weeks of preliminary research in the com­ ces. In addition scientist, but researchers, development agencies, or munity by a competent social one may also find useful the improved reliability of the data which fisheries offices, for information in church or government cen- are obtained will more than compensate histories, newspaper files, the costs of the research. It must be remem­ suses, local base records kept by wholesalers and retailers, bered, however, that key informants of fishing licenses or boat registra- their responses on their own experiences records not be tax records or records of equipment and that those experiences may tions, community. sales and imports. shared by other members of the It is important, therefore, to gather informa­ Attempts should be made to determine the tion from several key informants and to be reliability of all written records. Sometimes wary of information which is not confirmed inadequate data collection methods have by all those interviewed. Use of a sample been used, sometimes statistics were survey provides an opportunity to obtain inflated for political ends and sometimes quantitative data which can be used to esti­

77 Previous Page Blaz mate the variability of information of assessing taxes. If they are used to record obtained from key informants, all (or most) transactions they can provide continuous direct estimates of total catch 5.1.3 Census Surveys and effort and can he used as frame surveys of determine the total for designing random sample surveys A census is designed to most basic of different units in the individual landings. The fisherynumber suchof a varietyas landing sites, boats, fisher- receipts will give estimates of total landed revenue during a given perind of men, primary buyers or fishing gear. A cen- catch and of items landed fish which is not recorded and the sus is limited to the enumeration colected and counted at a number of receipts which are not which cri be observed are available. point in time. It requires periodic particular mix of units up-dating as the number and change. Census (or frame) surveys provide Landings may be recorded by weight, information which can be usf d to design volume or even number and are usually sample surveys. Sales receipts, if used reported by price category. The ex-vessel extensively, can be viewed as a continuous value reflects the landed weight, number or census of fishing activity. For a more com- volume times the price per unit weight, plete description of different types of census nur.ber or volume for each category and is surveys, the methods which can be used to summed over all categories to give the total conduct them, and subsequent catch assess- value of the fish which is sold to the dealer. ment surveys, see Bazigos (1974) and Since each receipt represents a single trans. Banerji (1974). action, direct estimates of nominal fishing effort are only possible if it can be assumed that each transaction represents a single shouldWhen conductingbe made to a distinguishcensus, some between attempt !) fishing trip by a particular kind ,of vessel. in use ver­ boats or gear which are actually sus those which are not, 2) part-time and The revenue represented by a receipt is the full-time fisherman, 3)dealers who buy and Tamount ruerepemen of money whiche d is yareigenerated is thewhen sell large and small volumes of fish, and 4) fish are sold to primary dealers and should important and not so important landing not be confused with the total value of the sites. Ideally, each landing site should be catch. Neither should the quantity of fish evaluated in terms of how much fish is landed and sold be confused with the quan­ landed, the principal species harvested, the tity which is caught at sea. Corrections types of gear used, how many boats nor- must be made in order to infer catch or mally land fish there, the number of buyers revenue data from landings data (see sec. who purchase fish, distances from markets, tions 4.3.4.2 and 5.2.2.1 (c)). communications and roads, facilities for holding fish, service facilities for boats and Additional information about the type of gear, etc. Similar information could also be vessel or gear used can sometimes he compiled for fishing communities and deduced by examining which species were include demographic information about the caught, and the name of the fisherman (or fishermen (ages, education, years of expe- vessel). In the case of mixed-gear, small­ rience, other occupations, etc.), their fami- scale fisheries; however, it may be difficult lies, and other members of the community.' to distinguish catch, effort and revenue sta­ In cases where it is not feasible to visit indi- tistics for individual types of gear. It is vidual sites byroad, data can be collected by sometimes helpful to know the location using aerial photographs or boats. For a where the fish was landed, and the naive of complete description of census methods, the the buyer who purchased it. reader is referred to the comprehensive manual by Bazigos (1974). Sales receipts can be designed so that much )epending 5.1.4 Sales Receipts more information is obtained. upon the dt-gr,, of cooperation which can be The quantity and ex-vessel value of fish expected from the dealers and the fisher­ which is sold to primary dealers are often men, a number of items - most of which recorded on sales receipts which are com- relate to nominal fishing effort - can he pleted when fish are weighed and the fisher- added. These will be very helpful in the men are paid. Receipts are sometimes filled designing of biological and economic sam. out by government officials for the purpose pie surveys. The most important items are 1)

I In practice, demographic surveys are extremely time consuming and expensiv', especially if there are many re-inoe communities which must he visited or if the fishermen do not live in permanent locations. Samplh surveys are recim. mended as a preferable method for obtaining sociocultural information.

78 a receipt number, 2) the gear-ty e used, 3) ual strata (e.g., landing sites within sepa­ the number of people fishing, and4) the trip rate geographic areas) and estimliting a length (days or hours). Only one of these pooled mean and variance for the entire (trip length) requires questioning the fisher- population from means and variances of man. Additional important items which individual strata. may be included but which may be more difficult to obtain are 6) the ;-one or area Three common types of sample surveys des­ fished. 7, the home port of the hoat, and 8) cribed in this guide are 1) landings surveys the name of the boat owner. for biological and economic data, 2) surveys of fishermen for sociocultural and economic Several incentives canhe to encourage data, and 3) catch surveys. Data collected the coopecration of the primary buyers. For when fish are landed (unloaded from a boat) example, receitts may be designed, printed ard sold to primary dealers are based on ands pplied to the primar lbuyers by the information gathered from examining the fishery department. In Costa Rica, multiple fish themselves (such as their size, species copy (3.4) receipts in serial order were pro- identificat;_n, and numbers), questions vided by the government for a one year asked of fishermen and the buyers (fishing period. The design, approved by several location, prices of fish, costs of operation) and observations of gear, time of day, etc. primary buyers, elicited a very high degree 2 of cooperatton. made by the interviewer. These data are 5.1.5 Sample Surveys used primarily for resource assessment pur­ poses and variable cost and returns analy­ are useful in situations sis. Sociocultural and economic fixed cost Sample surveys ran­ whert reliable estimates of total harvesting data are collected from interviews -of activity (catch. revenue, costs, effort;*-r domly selected fishermen. These interviews sociocultural characteristics of the partici- do not have to be conducted at the sarne time pants (dealers, fishermen) cannot 6e fish are sold since the anthropologist/soci­ obtained by complete enumeration without ologist is more interested in aspecta of spending considerable time or money. human behavior which affect the harvest. ing process, and the economist is interested Sample surveys are most often conducted in costs not related to the landing. Surveys according w tbh principle of randomization, of fishermen can, therefore, be conducted at a process by which observations, are ran- any time to generate economic, sociocultu­ clomly drawn from a "universe" of all possi- ral, and biological data, and be integrated. tile observations. The probability of selecting any single observation should be Catch surveys are designed to provide eI-ual to the probability of selecting any important information for resource assess­ other; the objective is to use sample data to ment purposes. Two major types of catch estimate the real value of properties (such as survey are 1) sampling aboard commercial means ar,1 variances) of the population of fishing vessels and 2) exploratory fishing all possible observations. The reliability of surveys. Because exploratory fishing sur­ sample data is evaluated statistically on the veys are usually carried out with gear such basis f the freedom from bias and the preci- as bottom trawls which can not be used in sion of the dalta. A biased estimate repre- many small-scale fishing habitats, and sm'nts some subset of the observations in the because trawl survey methods have been Iopulation more heavily than another (for thoroughly described in other sources, ra.her than small fish. or exploratory fishing techniquea will not be successfulexample., big fishermen fish rather than unsuc- considered in this guide (see section 3.2.4.4). cessful ones). Precision is a measure of the accuracy of the estimacA., i.e. the degree of Sample surveys should be designed so that similarity between thesample cstimate and they are relevant to location conditions. the corresponding population parameter. brief and easily administered.;, A relatively Biased estimates may he peecise. but they small preliminary survey of five or ten are not reliable. Precision increases as more respondents or landings will usually allow samples are taken, but increased sampling one to estimate how long it will take to com­ will not necessarily eliminate or plete each interview and to evaluate the e.ffortred'e, b!s, IPrecision can also be increased appropriateness of each item on the data val­ by dividing the sampling units into individ- form. Key infotmwits can often provide deliveries to primory dealers 'In wime rases. of raur*r.fioh may nat he delivered by heat. We will. ne'ert hele. refer it) all

if shellfish are not It nki etncu.ot i.. for e.xnmple. to ",I qurtionas which refer to shell fishing in the community h,"vi-MIa fit uid by the lxinl pIpuntiln.

79 uable advice during the pre-!iurvey trial with their occupation. A random sample period, can be drawn from such a list if it can be assumed that nearly everyone belongs to Experiences in numerous small-scale fish- the church. Lists of households compiled by ing communities have indicated that inter- local governments for taxation or other pur­ viewa in excess of haf-an-hour are not well poses can also bii useful. These lists must be received. If an interview taKes more than checked for accuracy before using them for half-an-hour to administer, it may be prefer- sampling rposes. able to divide it int two parts or to ask certain sets of questions only occasionally In cases where it is impossible to develop to randomly selected respondents. Use of a lists of fishermen without considcrable brief interview is especially important when effort, a quot i sampling procedure may be have just arrived from a fishing preferable. In this procedure, all landing fishermen fisher­ trip and do not want to be delayed. There is sites are visited and all small-scale limit to how long fish can be held for men (captains and crew) who are landing also a visit­ biological data collection before they must fish, working on their boats or simply or, sometimes, before they are sold ing the site during the time when the inter­ be iced be and taken away. viewer(s) is (are) at the site can interviewed. The interviewer should travel A number of factors must be considered from point to point starting at a different whcn designing random landings surveys site each morning so that interviews at the to obtain biological and economic data. In various sites will be distributed throughout general, the biologist focuses on the landing the day. This system eliminates possible as an opportunity to obtain information systeriatic bias caused by the landings of about the resource and the nature of its fish at any given site during a particular response to fishing pressure whereas the time of day. economist uses the landing as an opportun­ ity to obtain economic information about The randomization of the sample survey for the firm (individual fisherman o,"boat/gear economic fixed cost data is based upon the combination) and its production activity, major factor affecting the biological, physi­ Biologists may be more concerned with cal, economic and sociocultural nature of compiling catch and effort data for certain the production process - the boat/gear species or unit stocks and may therefore combination (firm type). The set of firm randon '?'e sampling effort over a slightly types will overlap with the set of fisherman set of landings - ones which are to be interviewed for sociocultural data just dffterent be m,.st likely to include the species of interest. as a large percentage of the landings will Econom;sts, who are alo concerned with common to post-trip interviews and land­ catch and effort data, will randomize their ings surveys, especially those which are sampling effort over all landings. The objec- designed to produce catch and effort data. is t[" same in either case - to tive, however, Activities design a sampling procedure which will dis- 5.1.6 Research tribute sampling effort in proportion to the Research may be necessary to provide some actual distribution of a sampling unit as of the background data necessary to des­ determined from sales receipts or a census cribe resource habitats and ecosystems (see survey.4 To the degree that biologists and economists are interested in the same varia- section 4.1.1) and for providing the vital sta­ bles, the same randomization procedure can tistics necessary for assessing and manag­ and implement- ing resources by means of dynamic pool be followed when designing research activities and ana­ ing a random landings survey (see Chapter models. The lyses which are referred to in this chapter VD). generate estimates of gruwth and mortality fairly simple order to conduct ,,ciocultural surveys of rates and have been limited to In and recapt-ure stu­ fishermen or owners of fishing equipment techniques such as tag respondents must be selected dies, the analysis of growth rings on scales and vessels, of size fre­ randomly from lists of names. Lists of and otoliths, and the analysis owners of equipment can quency data. These techniques may require fishermen and/or and be obtained from census data. For example, some degree ofexpertise and experience countries, the local parish are more time consuming than data collec­ in some Catholic to generate a keeps a list of all church members along tion efforts which are designed

siteor sampling unit forms the basis for the sampling process: it may be a landing site, an individual boat at a landing A units could be an individual landing i'self. A frame survey is a l:st of all sampling units from which a set of sitnpling selected.

80 time series of catch and effort data. In situa- any interview data. Supervisors should tions where rc'Nble catch and effort data examine initial interviews for inconsistent cannot be easily obtained and exploratory and/or inadequate data. Problems encoun­ fishing surveys are not feasible, these tered in interviewing should be discussed research activities may be the best source of with the interviewers so they will be aware resource assessment data. that conducting good interviews is a diffi­ cult process that must be learned, and that a 5.1.7 The Interviews good interviewer is a valuable person. If the is the interviewers are advised that it is quite com­ OfaOne pointthatnt that everyoe mustst be emanemphasized siod as te mon for. some informants to either mislead fact that not everyone can function as an them or refuse to talk to them, they will be ef.'ective interviewer. The selection, train- more likely to mention these problems dur­ ing and supervision of interviewers is a cru- ing the review sessions, and the supervisor cial aspect of any data collection program. can then help develop strategies to minim­ Decisions based upon information gener- ize or eliminate these problems. ated from inadequate data or data which have been improperly collected can have Interviewers need to be supervised. If the negative impacts on the entire fishery. It is interview results are analyzed during a pre­ therefore important that careful attention liminary trial period as suggested above, be paid to all aspects of the interview the interviewers will know that their work process. will be reviewed. This kind of supervision by mature, helps discourage the falsification of data. The best data are collected visits at interview­ well-motivated interviewers Periodic, unannounced po sed, and ing sites or visits with informants after whowoi cananestablishel somesm edereofrdegree of rapport interviews have been completed are key ele­ with the respondents. It is therefore neces­ sary that the interviewer respect the ments of supervisingactivity. respondents and be sensitive to social or 5.2 BIOLOGICAL/ECOLOGICAL cultural barriers which impede the collec- DATA COLLECTION METHODS tion of reliable data. Frequently, interview­ ers project an image of superiority which 5.2.1 Habitat/Ecosystem Data makes it difficult, if not impossible, to estab­ lish the type of relationship which makes it 5.2.1.1 Biological /Ecological possible to obtain reliable data. If the Inventory respondent feels that the interviewer is treating him as an inferior of if he feels that Much of the data needed for an inventory of the interviewer considers thatdata to be of the more important exploited and underutil­ little value, he will not make an effort to ized species can often be obtained from pub­ provide good information, and in some lished reports for these same species or for cases may even mislead the interviewer, other closely related species. Since the col­ lection of new biological/ecological data When training interviewers, it is important requires laboratory and field studies which to provide a good description of the nature can be expensive and time-consuming, first and significance of the data which will be priority should be given to the retrieval of collected. The interviewer should under- information from published sources." stand how the data will be used to make policy decisions which will affect the wel- The first step which must be taken in pre­ fare of many people. If the interviewer paring an inventory is an identification of understands the importance of the inter, the species which are harvested. Species view, he can pass the information along to identification must be based on accepted his informants. Th6 sharing of this infor- international nonmenclature, i.e. scientific mation can result in the more willing coop- names rather than common names. Identi­ eration of the informants, fication requires the use of taxonomic keys and must be performed by competent biolo­ Many of the problems encountered by inter- gists who are familiar with the technical viewers occur during the first few inter- terms used in describing the morphological views; hence, it is necessary to continue characteristics of fish. Once a list of scien­ training sessions into the first stages of the tific names has been compiled, it should be interviewing process or to conduct pre- cross-referenced with a list of common survey trial interviews before compiling names thus facilitating the collection of

-1An excellent series of resource inventories have been published recently by the United Nations Food ant Agriculture Organization (see Fischer, 1978 for example).

81 species-specific biological and economic Methodologies which should be followed in data from fishermen and dealers. Ideally, a stomach contents analyses were reviewed reference collection of preserved fish should by Hyslop (1980). Once the ecological roles be maintained for identification purposes. of individual species within the ecosystem are known, it may be possible to predict that the harvest of a particular prey species will 5.2.1.2 Extent of Fishing Grounds reduce the growth and/or survival of a pre­ pre­ of grounds where fish are or dator while the harvest of a particular An inventory the growth harvested can be compiled by exa- dator species will increase could be prey species." mining bathymetric navigational charts. and/or survival of certain Calculations of area for fishing grounds on Since the feeding habits of fishes depend on continental shelves, reefs or offshore banks the relative availability of different prey can be made with a planimeter. Some infor- traive avl t y ofediferntpe mation on bottom type is usually printed on organisms and the feeding behavior of the these charts as well. Information concern- predator at any particular point in time, which are used to harvest attention should focus on the collection of ing the gear types simply indicate the certain species or groups of species in cer- qualitative data which be obtained from fisher- major sources of food which the principal tain habitats can depend key informants, commercial species in the fishery men or other upon at different times of year and, if appli­ Production Rates and Ecosys- cable, at different ages. This information 5.2.1.3 recording tern Dynamics can be collected by observing and stomach contents or, frequently, by refer­ As much information as possible should be ring to published reports. Many of the illus­ obtained from existing sources since trated field guides to the fishes of major research is time consuming and expensive, regions of the world's oceans include infor­ A rough evaluation of the relative impor- mation on major food organisms. tance of different sources of primary produc­ tion can be based on an examination of 5.2.1.4 Hydrographic Data charts which indicate coastal navigational rivers depth, bottom types, the extent of reefs and Data on freshwater runoff for major mangrove swamps and the number and size are often collected Fznd published by inter­ of rivers, bays and estuaries along the coast. national organizations or national agencies Reasonable estimates of average annual responsible for water resource manage­ primary production can sometimes be ment. Information on currents and tides is obtained either from published reports (see regularly published for use in coastal navi­ FAO, 1972, for example) or by direct mea- gation (see U.S. Department of Commerce). surement. Several guides to field techniques Additional information on currents and are also available (Vollenweider, 1969; U.S. upwellings is sometimes available in publi­ National Academy of Sciences, 1969). It is cations of oceanographic research. not difficult to measure primary production - the problem is that production rates vary 5.2.1.5 Other Physical/Chemical Data considerably over time and from place to place. Thus, reliable estimates require use of Since the development of empirical models a standardized technique and multiple which relate yield to some set of environ­ measurements. mental variables would be largely an exploratory exercise, it is impossible to pre­ Feeding habits studies require the examina- dict what kinds of data would be required. tion of fish stomachs collected during Collection of new data requires that envir­ research surveys or from fishing vessels. onmental variables be monitored for at Stomach content analysis should be quan- least a year. Efforts should first be made to titative at least to the extent of determining utilize data such as productivity rates, areas the relative frequency of occurrence of of continental shelf within a certain depth major prey taxa (for example, polychaetes, range, temperature or salinity which are fish or crustaceans). The more specific the already available either to generate empiri­ taxonomic identifications are, the more cal yield functions or to provide g-nral readily can the detailed feeding relation- background information which describes ships within the ecosystem be defined, habitat and ecosystem characteristics.

"In reality, predator-prey dynamics are more complicated than this.Some predators, for instance, are much morespecific in their dietary preferences than others while others easily switch from one prey to another, depending on the relative abundance of prey organisms.

82 5.2.2 Stock Assessment Data fish and the species con ,sition ofdifferent price categories. 5.2.2.1 Catch and Effort Data can be Sales Receipts Weight/volume conversions 5.2.2.1(a) obtained fairly easily by calculating the are usually recorded on even the average weight of a series of samples of Data which conver­ crudest type of sales receipt when fish are known volume; weight/number from the fishermen include 1) sions depend on the size distribution of the purchased price and how much fish was sold, 2) when and where fish which are sold for a given the transactioi, took place, 3) the nqmes of therefore must be established for individual the buyer and seller, 4) the unit price and sizes or size groups. These conversion fac­ landed value of the fish which was sold for tors should be estimated from commercial individual price categories, and 5) the total catch samplea collected at different times of value of the catch which is sold to the dealer. year since the average weight of a given data which may be recorded volume or number of fish of the same length7 Additional course of the year. include the type of vessel or gear which was may change during the used, the riumber of fishermen who were Similarly, species composition data should aboard the vessel (or who caught the fish, if be obtained from landings surveys which no vessel was used) and the general location are repeated periodically dring an entire where the fish was caught. Some of this year since the species mix in each price cate­ information can sometimes be deduced by gory can be expected to change as the avail­ knowing when and where the fish was sold. ability of different species and their marhet All of these data, except price and value value changes. For fisheries which harvest information, are useful for stock assess- only one or a few species, landed weight ment purposes. data can be utilized directly for stock assess­ ment purposes without landings survey The quantity of fish which is sold during estimates of species composition as long a­ any single transaction may be recorded by each species has a unique unit price. Since weight, volume or even numbers. Since most tropical small-scale fisheries harvest a most resource assessment techniques are variety of species, some random sampling based en biomass estimates, landings data of landings will be necessary to even get a which is reported in units of volume or rough estimate of species composition. have to be converted into units of numbers species weight (usually kilograms). In addition, the Once landings data are available by quantity of landed fish is usually recorded or species groups on a weight basis, they by price categories which include all species should be converted to catch data by esti­ and sizes of fish ofthe same price at the time mating the quantities which are captured, the transaction was made. The species but not sold to primary dealers. This can be which are included in any given price cate- accomplished by estimating post-harvest gory at any particular time change as a losses which take place at sea (see section result of price changes for individual spe- 5.2.2.1 [c]) or after the fish are landed (sec­ cies (or sizes). Because price can be a func- tion 5.2.2.1 [b]). The most significant kinds tion of size as well as species, it is not of post-harvest loss in tropical small-scale uncommon for smaller individuals of the fisheries are likely to be spoilage and fish same species to be included in one price cate- which never reaches the primary dealer gory and larger individuals in another. because it is taken home and consumed by or These groupings usually have no biological the crew and their families, given away, i.e. they include fish of differ- sold to someone else. Another important significance, catch data ent species which share no common biologi- source of error when compiling landings surveys) is cal or ecological characteristic and which from sales receipts (or be considered to belong to fish which the fishermen are given or which cannot, therefore, vessels such as a single unit stock (see section 3.2.2). Thus, they purchase from larger useful stock assessment shrimp trawlers. In practice, it may only be in order to extract of sales receipts, conversions must possible to make rough approximations data from reported on be mude to obtain landed weight data by how much fish is caught, but not species groups. These conver- sales receipts (or reported, but not caught). species or per­ sions require some information on the aver- Stock assessements are frequently age weight of a given volume or number of formed with uncorrected landings data.

for instance, r the fact 7 These changes may reflect the development of reproductive tissue during the spawning season, o volume or that fish are more likely to have ful stomachs when food is plentiful. Estimates oflanded catch (weight) from number data are much more straightforward if fish are gutted before sale.

83 Effort estimates obtained from sales Standardization procedures may be devel­ receipts are usually crudely expressed in oped only when it is known what technolog­ terms of the number of trips if it can be ical factors affect capture efficiency and assumed that each transaction represents when information exists which reveals his­ the sale of fish caught during a single fish- torical changes which have taken place in ing trip (see section 5.1.3). Again, informa- these factors over time., during landings surveys may tion obtained in to identify which kinds of fishing activ- Despite the problems which are involved help from sales ity - defined in terms of gear type, species obtaining catch and effort data fishing location or time of year - receipts which are useful for stock assess­ captured, offer a unique could be expected to violate this assumption ment purposes, receipts data necessary to estimate the opportunity to directly estimate total catch and provide reli­ average number of trips associated with a and effort. If the data are complete and single transaction or the average number of able, there is no need to rely on extrapola­ hours or days associated with a single trip. tions of total catch and effort based on Such refinements in fishing effort estimates landings surveys which may not be truly are extremely important for making preli- random and which may only sample a minary or quantitative stock assessments small proportion of all landings. Implemen­ and permit a much greater utilization of tation of a comprehensive procedure for col­ sales receipt data. lecting standardized sales receipts should therefore be the first priority of any effort to Catch and effort data obtained from sales monitor the activity of a small-scale fishery receipts should be compiled, as much as pos- which delivers the majority of its harvest to sible, for individual types of fisheries, dealers who routinely keep some kind of Although stock assessments may, in fact, records or who can be encouraged to do so. be performed by combining catch and effort In situations where sales receipts can not be data for d Cerent gear and vessel types, it is relied on to provide reliable estimates of equally possible that the resources har- total catch and effort, they must be esti­ vested by the gill net fishery, for instance, mated from landings surveys. will be evaluated independently of the resources harvested by the handline 6.2.2.1(b) Landings Surveys fishery. For this reason, therefore, any addi­ tional information which reveals what type Landings surveys provide both complemen­ of gear or vessel was used is e'tremely valu- tary information for improving catch and able. By th3 same token, inforrration which effort data obtained from sales receipts and can be used to determine even the general primary data for stock nssessment purposes location of the fishing grounds where the when sales receipts are not available. They catch was harvested is very useful for unit are conducted by trained observers who stock assesfments. Such information may record data as fish are delivered for sale to a be deduced by knowing where Lhe transac- primary buyer. Deliveries are usually made tion took place, an item which is usually when the fish is "landed" by the fishermen, recorded on the receipt. i.e. when the fishing boats are unloaded at the landing site and the fish is weighed and Fi, .Aly, kncwing the date when each trans- sold. Data are recorded by interviewing the action takes place is crucial since catch and fishermen and the buyer, weighing all the effort statisirhs must be compiled for certain fish of any given species (or species group) periods of time (months, years) in order to which are landed, and making observa­ reveal trends in catch per unit effort (an tions. The need to identify each fish requires indicator of resource availability) over time that landings surveys be conducted by peo­ as effort changes. The compilation of a time ple who can easily recognize different series of annual catch and effort data species. requires that changes in the capture effi­ ciency of a given fishing technique be incor- Ideally, the landed weight should b;. porated into annual estimates of fishing recorded by species within price cei ggories. effort. Techniques for standardizing effort This rule should certainly be foliwed for have been developed for some of the more the principal species in the fishery, i.e. those common large-scale types of fishing opera- which account for the bulk of the landed tions, but not for most small-scale fisheries, catch and its value. If thete are a great

"For example, if monofilament gill nets are known to be twice as effective as multifilament nets, standardization would require information on the relative number of nets used in the fishery which were made from both naterials during each year when effort data were available.

84 many species within any given price cate- ing the entire area which is fished into fairly gory which are morphologically similar and large areas. The compilation of catch and difficult to distinguish from each other, effort data by location is more critical when those which are caught in small quantities effort is dispersed over a large area or when can be weighed together and identified by unit stocks occupy small areas. genus or even family, rather than species. The weighing process can be time consum- A great deal of useful information can also ing if large numbers of fish are landed or if be obtained by simply counting the number they are not easy to identify; care should be of fishermen aboard th2 vessel and by not­ taken not to impede the buyer's efforts to ing the type of vessel and gear, how much of sell the fish or to process them in order to the landed catch is sold to the dealer and prevent spoilage. In these situations, data how much is given away, taken home by the collected rapidly by weighing a ran- fishermen, or discarded because it has can be in domly selected subsample of each price cate- spoiled. An important clue which helps careful not to bias the sample by identifying the fishing location is the direc­ gory, being it selecting large fish or particular species in tior which the vessel comes from as preference to others (see Gulland, 1966 for a approaches the landi.g site. of techniques for sampling large description data quantities of fish). A significant problem which impedes collection when using sales receipts or a which are Valuable data are also obtained from the landing survey is that the fish The most important data which landed on any single occasion sometimes fishermen. trip nor the fishermen provide are estimates of fish- are not caught on a single fishing Effort estimates require some with a single type of gear nor even by the ing effort. The knowledge of the amount of time that was fishermen who are aboard the vessel. devoted to catching the fish which are being vessel may simply be delivering fish which a description of certain character- were caught by a number of different fisher­ sold and trips. istics which affect the capture efficiency of men on a number of different fishing gear which was used. A .ummary of On such occasions, valid catch and effort the if each fishing effort estimates for some common data may sometimes be estimated fishing gears is presented in fisherman's catch is weighed and sold small-scale are Table 1 (Chapter IV). Fishing time is usu- separately and if the fishermen who ally estimated as the time the gear was actu- delivering the fish can recall approximately the water. Some important fishing how much time each fisherman spent fish­ ally in he power estimates are net size (length x ing, where he was fishing and what gear number of fishermen, and numbers used. In any case, it is up to the interviewer width), possible of gear units (pots, hooks) fished during a to record as much information as it is. particular trip. These data are obtained by and to decide how reliable the fishermen several simple ques­ asking total number tions such as "how long did you fish," Since only a proportion of the "what gear did you u-.," "how many (nets, of landings (or boats) is sampled, landings f'--the do you use," and "what is surveys must be designed to account pots, lines, hooks) of fishing of your gear?" If the fishermen are spatial and temporal diitribution the size the full range willing, these data can often be obtained (or effort and to randomly sample by examining the gear when the of fishing activities which affect the catch verified) stocks. Factors fish are landed. Examples of questions rates of individual unit were included on a landings survey which should be considered when designing which stock assess­ form which was designed to generate catch random landings surveys for data from a small-scale fishery ment purposes include the fishing location, and effort vessel types which was conducted primarily with gill the landing site, gear and and longlines is presented which are used in the harvesting process nets, handlines as the time of in Appendix C. and time considerations such year and the state of the monthly or annual factors can affectthe Additional information concerning the tidal cycle. All of these which was abundance and availability of the resources location of the fishing ground(s) in any single trip should and the relative success of the fishermen (were) visited during which during landings surveys capturing them. Landings surveys also be collected effort to a it can be important for defining which allocate more or less sampling since gear type, time of year resources or unit stocks were harvested, particular location, do not willingly reveal or month can easily produce biased esti­ Although fishermen effort for all loca­ locations of favorite fishing grounds, mates of mean catch and exact of year or months. this information can be obtained by divid- tions, gear types, times

85 In order to allocate sampling effort ran- points in time.9 The size of these fishing domly, a primary sampling unit must be locations would require some knowledge of defined and an initial census (or frame sur- individual unit stocks and the willingness vey) of sampling units carried out. The sam- of the fishermen to divulge information pling unit snould represent an element of about specific fishing grounds when they fishing activity which can be easily quanti- are interviewed. fled such as individual vessels, landings, or landings sites. A survey could be designed In practice, the logistical difficulties to randomly sample some proportion of all involved in conducting adequate frame sur­ landings for different gear types. It would veys and the lack of sufficient biological first require a census of the number of land- information about the stocks which are ings by gear type in all landmg sites during harvested by most small-scale fisheries pre­ some previous period of time so that sam- clude the definition of sampling units which pling could be allocated representatively incorporate fishing location as a variable. between landing sites for each gear type. Catch and effort data obtained from land­ Effective landings surveys can be designed ings surveys which are randomized to if continuous information on total catch account for the distribution of boats (or and effort by gear type, landing site, and fishermen) and gear types by landing ite time of year is available from sales receipts. should, hcwever, still be compiled by fish­ ing location. In the absence of adequate bio­ Frame surveys which are conducted at a logical information, the definition of given point in time must be periodically appropriate locations in most cases must.be updated to account for changes which take made on a purely geographical basis, place in the fishery. These changes could although some attempt should be made to include ar. incre'se or decrease in the define locations in which most of the fishing number of fishermen, vessels or units of effort is conducted by boats which deliver gear or a shift in the distribution of landings their catch to the same landing sites. by gear type among landing sites. Observa­ tions made during the landings survey itself Landings surveys should be randomized to can provide current information useful for account for the number of boats which updating static frame surveys and adjust- deliver fish to individual landing sites. Cen­ ing sampling strategies. In this mode, the sus surveys ofthe numbers of boats in use at sampling procedure is continually each site or sales receipts which document improved as new information becomes numbers of landings made at each site per available and as conditions in the fishery unit time can be used as frame surveys for change. drawing random samples. This information may be qualitative or quantitative (see One of the most difficult variables to Banerji, 1974); the more quantitative it is, account for in designing a random sample the more precisely can the sample statistics survey of landings is the fishing location. (e.g. mean catch per boat) be estimated. Since samples are not obtained while boats Data collection procedures which generate are on the fishing grounds and since the catch and effort statistics for resource distribution of landings in different sites assessment purposes can be improved even along the coast does not necessarily reflect further if sampling effort is allocated pro­ the distribution of effort between different portionally among boats (or landings) fishing locations, the best solution would be which use the same gear type since the type to design a procedure for randomly sam- of gear which is used determines which spe­ pling landings in which the same propor- cies or stocks of fish are harvested. The allo­ tion of all the effort expended in each cation of sampling effort by gear type and location per unit time is represented in each landing site will also greatly improve the sampling period. In other words, fishing representation of different fishing locations location should ideally be included in the in the sampling scheme since the fish har­ definition of the sampling unit. This vested on certain grounds tend to primarily approach is not feasible unless repeated be caught with particular gear types and frame surveys could be performed which landed in particular locations. One of the would permit the enumeration of vessels problems which impedes randomization by fishing in specific locations at particular gear type is that many boats use more than

9These surveys would have to be repeated frequently in order to account for changes in the spatial distribution of effort which take place over a very short time scale (months, weeks, oreven days, depending on how specifically the locations are defined). The only practical method for conducting such surveys may be aerial photography; such a technique would be limited to day time fishing only.

86 course of the year or include seasonal changes in temperature, one gear type over the as a single fishing trip. When this currents, and the amount of river runoff even during changes in is the case, "mixed gear" elements must be well as short-term (monthly) included in the sampling scheme. By the tidal currents. same token, if two or more landing sites are Most of these events recur with a predicta­ close together and the same boats are just as ble frequency and can be factored into a likely to land fish at either site, the sam- random sampling strategy; many of them pling unit should include a cluster of land- recur on an annual basis, but some recur ing sites as a single data collection point, over much shorter time periods (weeks or in tidal amplitude and considerations are the months). Changes Other imporant currents, for example, take place over a 28­ duration and frequency of the landings sur- influence to a large day period and can exert a strong vey.tSampling strategy depends and feeding behavior toof vey.ntextent on whetherin tr or not nottotal to catcha and onfish the (and swimming therefore their susceptibility capture with certain gear types),the perfor­ effort can be reliably estimated for the the dirtr­ types canue of certain gear types), entire fishery (and for individual gear effort. Given receipts. If sales bution and amount of fishing and unit stocks) from sales tIose short-term effects relied upon to provide the importance of receipts car not be long-termn (e.g. annual) sam­ ­ either because a large on lartdings, this infor.ation ffort should ideally be distributed of the fish which is caught is not piing percentage over different fishing seasons and short sold to primary dealers or because a signifi- should is term (e.g. monthly) sampling effort cant proportion of the landed catch which actual changes - then land- be distributed so as to reflect sold to dealers is not recorded and type of landings (as be carried out on a con- in the number ings surveys should by gear and vessel type) in individ­ order to provide primary defined tinual basis in sites a a function of factors such as assessment purposes. One ual data for stock tidal amplitude. Even the time of day (or of a continual survey is that can be impor­ advantage of night) when fish are landed sampling is spread out over a long period tyne d an i ng since n t if, f ih ar time and need not be very frequent for a given gear type and landing short term precision of catch arid effort esti- tant if, lng ermsite, landings during one time of day repre­ mate iporantasisnotas sent a different set of activities (fishing loca­at trends.matres isIf notsales as receiptsimportant can asbe used,long termthen tions, for example) than landings time of day. to be conducted peri- another landings surveys need changes in order to determine odically often which may take place in the species compo- There are a number of problems which sition of landings, the average number of make it impossible to completely follow a hours or days spent fishing during each randomized sampling procedure even when fishing trip and the characteristics of the sufficient information is available to permit gear which were used. Periodic surveys its design. The spatial distribution of sam­ could be carried out, for example, by sam- pling effort among individual landing sites pling frequently during short periods of requires either that interviewers remain sta­ time (such as fishing seasons) when pat- tionary in a given site long enough to obtain fishing activity can be expected to the necessary samples or that a sampling terns of inter­ remain more or less constant. These periods schedule be devised which permits sampling should be repeated viewers to move from one site to the next of intensive samples in a often enough and last long enough to detect without completing all their in resource availability and har- given site during a particular time of year, changes of day. Since many vesting activity which take place over time. time of month, or time small-scale fishe-ies are characterized by sites (with or without sales large numbers of remote landing In either case certain times it is important to know something which may be inaccessible at receipts), of random sampling about the temporal variability in resource of year, the design and availability in order to procedures which can in fact be followed abundance of time, design random landings surveys which are without considerable expenditures particular periods when manpower, and funds is a challenge (see not biased towards must catch rates for particular species may be Chapter VI). Another problem which or below average. Important biotic be confronted is the conflict between office above sche­ factors which affect resource availability working schedules and fishermen's and the distribution of fishing effort include dules. It may be impossible to sample land­ seasonal migratory patterns, spawning ings at night, early in the morning, or on the behavior, and the abundance and distribu- weekends unless non-government person­ tion of food organisms. Abiotic factors nel are hired to do so.

87 5.2.2.1(c) Sea Sampling we:ight estimates rather than maximum length estimates are used in dynamic pool Periodic observations made aboard small- yield models. scale fishing boats can provide some basis for estimating the types (species), sizes, and quantities of fish which are caught, but not landed.' These observations should .2.2.2(al) Tag and Recapture Studies include fish which are discarded at sea and Growth increments during specified time fish which may be used for bait, consumed periods can be calculated for individual fish aboard the boat, spoiled, or sold before the which are measured at tagging and again at catch is delivered to the primary dealer. If recapture after a known period at liberty. In fish are gutted before they are landed, an this case, the growth coefficient (K) is esti­ estimate of the percent weight loss after gut- mated from the slope of the linear regres­ ting is necessary. If fish are discarded, some sion when Y equals the change in length estimate of the proportion which do not sur- during a given (constant) time interval and vive should be made. If other fisheries (for X equals length at tagging; the length at example, trawl fisheries) are harvesting which the change in length is zero (the X ­ and discarding significant quantities of the intercept) equals L .When growth data same species wAich are captured by the from all fish returned during any number of small-scale fishery, some time should be short time intervals are used, K may be esti­ spent aboard these vessels as well. mated from the slope and L . from the X­ intercept when the increment per unit time Sea sampling data can not be expected to is plotted against initial length. For equal provide very reliable estimates of how much time periodls, when Y equals the length at catch recapture and X equals initial length, K is ratesfish is are caught so variable but not and landed since thesince amount estimated from the slope and L . from the each trip is small. Sea intersection of the linear regression with a of fish caught during lenth should only be carried out occa- 450 line representing no change in sampling The parameter t. sionally and the results should be comple- during the time interval. with landings survey data and data mented any while conducting sociocultural can be estimated from the VBGF for collected known length at age once L w and K have surveys of fishermen. been estimated. For longer, unequal time 5.2.2.2 Vital Statistics intervals between tagging and recapture, other methods are necessary. For a com­ 5.2.2.2(a) Growth plete description of analyticai tecbniques used in tag and recapture studies see Jones A number of mathematical functions are (1976). used to express the growth history of fish. The most common is the von-Bertalanffy One of the most persistent problems asso­ growth function (VBGF): • ciated with estimating growth from tagging It = L.(1.-e-k(tc-to)) data is measurement error. Errors as small as ± 5 mm or less can be significant when This function may either be fitted directly to time intervals between tagging and recap­ length (or weight) at age data or after ture are short. When large numbers of mathematical transformations to various returns are available, data obtained from linear regression formulae (see Ricker, 1975 fish recovered during an initial phase of or Gulland 1969). Size at age data may be minimum growth can be eliminated. This is obtained from fish of known absolute or rel- also the period of time when the effects of ative ages using a variety of techniques tagging injury on growth will be most pro­ such as 1)tag and recapture studies, 2) anal- nounced. Attempts should be made to actu­ ysis of growth rings on scales and other ally recover all recaptured fish for precise hard parts, and 3) the analysis of incre- measurement. Also, data obtained from fish ments in size frequency data collected over a which are iced or frozen should be corrected period of time. Some methods allow for the to account for shrinkage. Care should also simultaneous estimation of maximum be taken to use the same measurement (for length (L ) based on length at age data. length, standard length or total length) for L may be converted to maximum weight each species. A complete description of (W. ) if the length/weight relationship for materials and methods used in marking stu­ the species in question is known. Maximum dies is given by Jones (1979).

See (Gulland(1966) for a description of data collection techniques which npply to sea sampling.

88 5.2.2.2(a2) Analysis of Hard Parts fish are landed for delivery to buyers. When large quantities are available, sub-samples It is known that the growth of fish scales, should be randomly drawn to avoid biasing fin rays, otoliths or vertebrae is propor- the measurements toward either large or tional to the growth of the entire animal. If small fish (see Gulland, 1966). Lengths are it can be confirmed that growth rings laid more easily and precisely recorded than down by cyclical interruptions in the ani- weights and may be converted to weights mal's growth are caused by natural events once empirical length-weight relationships (such as low temperatures, low salinities or are known. Length-weight relationships for slowed growth during spawning) which individual species may be determined from recur with a known periodicity, the growth length and weight data collected for several history of individual fish can be deduced hundred fish selected randomly over the from the length of the fish at capture, the course of a year. Care must be taken to com­ distance from the focus of the scale (for pile data separately for male and female example) to its edge and to each growth fish, since growth rates may vary by sex, ring. This method requires that each ring and to include fish with both ripe and spent correspond to an absolute age. Once an age- gonadal tissue and fiRl and empty length key has been established by examin- stomachs. ing scales (or otoliths, etc.) from a number of fish, a growth function is fit to the data to Once primary size data have been collected, estimate parameters. Graphical methods they should be compiled by equal length or described in the previous section are weight intervals. Ten to fifteen intervals are applicable, normally sufficient to permit a visual inter­ This methodology does not apply as well to pretation of component size groups. All size tropiclhfishd s g eiessincegris not as data must be recorded by fishing locations, tropical fish species since growth is not as gear and time of year since all of these fac­ predictably cyclical as its for temperate tors may affect the sizes of fish captured. There is, however, no need to randomly produced,zone species. but Annular may be ringscaused are by sometimes variable select certain landings. Instead, attention cuon landings whion such as strong wind and seld natural events focus on those landings which rainfall as well as seasonal changes should heavy supply the greatest number of fish. in temperature, salinity and spawning activity. For growth analysis, understand- Growth rates can be estimated from the pro­ ing the causes of cyclical growth are not as gression of modal or mean lengths of indi­ important as the fact that ring formation vidual size groups which recur in size must correspond to some predictable, recur- frequency distributions collected over time ring natural event. In general, the analysis by plotting the natural logarithm of the of hard parts for growth information is a quantity L . - It, (y), versus It, (x), when It time-consuming process which requires a equals the modal or mean length for a given great deal of judgement and practical expe- size group at time t. The growth rate during rience. For an excellent review of age and the interval between samples is estimated growth studies as they auply to tropical fish from the slope of the regression; the "best" species, the reader is referred to Brothers estimate of L . is that which produces the (1980). most linear array o'data points. Estimates by species is not of K and L . are thus obtained The growth of invertebrate to fit the Crustaceans, for repeated trial-and-error attempts quite so straightforward. to the data when L . is only increase in size when they "best" regression example, to vary. Once L . is known, t., can molt (shed their hard exoskelton); g-owth allowed molts be estimated from the y-intercept (which rates are a function of the number of age at each molt. equals kt).Information on the absolute and the increase in size is not required. in an interrupted, stepwise of individual size groups Growth prcceeds of individual size groups Mollusks which are subjected to Since the growth fashion. be followed through time, it is easier to seasonal changes in environmental factors nust produce identify the same size group in successive such as temperature and salinity samples is in their shells which may data sets if the interval between growth rings results are obtained from be used to estimate growth rates. minimal. The best sometimes continuous monthly (or even weekly) size 5.2.2.2(a3) Size Frequency Data frequency data.

Size frequency data can be obtained by The use of size frequency data to estimate measuring individual fish either aboard growth paraieters is difficult when 1)there fishing vessels or a research vessel or when is excessive overlap between adjacent size

89 groups, 2) sample sizes are small, and 3) estimated by subtraction. Assuming that recruitment of juvenile fish to the exploita- tag returns are being reported by fishermen ble population is continuous or nearly so. and that fishing effort does not remain con­ When overlap is extreme, component size stant with time, more reliable estimates are groups are not visually obvious and graphi- obtained when the number ofreturns in suc­ cal and mathematical procedures for esti- cessive tine intervals is recorded on a per mating the mean lengths of individual size unit effort basis, especially if fishing effort groups produce unacceptable results even varies significantly over the course of the when sample sizes are large (Stevenson, year. 1980). There are many sources of error inherent in The number of size groups and the degree of this technique which can be grouped into overlap between them increases when there three categories depending on whether they is more than one annual spawning season, affect F only, Z only or both F and Z (see when spawning is prolonged and when Gulland, 1969; Jones 1976). Natural mortal­ male and female fish grow at different rates. ity estimates may or may not be affected, When recruitment is continuous, size incre- depending on the degree to which F and M ments in time series data can not be are affected. Some errors can be corrected detected. These phenomena are common in during the experiment, others cannot. In tropical fish species. Also, the best results warm tropical waters infections due to tag­ are obtained when fish are captured by gear ging are more likely to cause problems, and which is not size selective (seines, traps) fish are more susceptible to injury during since selective gear (gill nets) severely res- handling. Also, the non-random distribu­ tricts the range of sizes captured. tion of tagged fish which are returned to the population is a major problem which may for demersal tropical 5.2.2.2(b) Mortality speciesbe more with significant limited geographic ranges. require estimates of Dynamic pool models instantaneous natural morality (M) and Tagging experiments conducted with spe­ fishing mortality (F). I Together, these two cies which are captured by small-scale estimates are summed to estimate total fisheries often require the capture of large instantaneous mortality (Z). The time inter- numbers of fish in locations which are not val of interest is usually one year in which accessible to efficient harvesting tech­ case mortality is expressed as an annual niques, for example, in mangroves and on instantaneous rate. The first step in esti- coral reefs. Since only a 5- 10% recovery rate mating M and F is usually to estimate Z can be expected it is necessary to tag and since total mortality is more easily deter- release several thousand fish for each spe­ mined. In the special case of an unexploited cies which is being studied. Furthermore, population, all mortality can be attributed most tagging experiments rely on fisher­ to natural causes and Z = M. Three of the men for returns. In small-scale fisheries, most common methods for estimating mor- tagged fish are often landed in remote land­ tality are 1)tag and recapture studies, 2)the ing sites which frequently do not have any analysis of size frequency data and 3) means of rapid communication with major changes in total mortality aid fishing population centers where fishery offices effort. tend to be located. Moreover, fishermen may be suspicious of government research pro­ 5.2.2.2(bl) Tag and Recapture Studies grams. Successful recovery efforts therefore require extensive publicity programs and If tag returns are grouped in equal time efficient methods for recovering tagged fish intervals, the plot of Y (the natural loga- and paying rewards. (It will, of course, not rithm of numbers returned during cach time be necessary to recover the tagged fish interval) versus t time intervals is linear themselves unless growth measurements with slope equal to -Zt. Furthermore, know- are being made). If enough tagged fish are ing the total number of tagged fish and the released the percentage of non-reported re­ numbers of tagged fish caught in successive coveries can sometimes be estimated by intervals, F can be calculatcd for any inter- comparing the recovery rate from a group of val or F can be calculated from the Y- cooperative fishermen. (or boats) with the intercept. Once Z and F are known, M is recovery rate for all fishermen (or boats).

"Instantaneous rates of change equal the natural logarithm of the absolute finite rate. In the case of mortality rates, the absolute finite rate equals the percentage of fish which die during some time interval. Finite rates cannot be summed whereas instantaneous rates can.

90 5.2.2.2(b2) Size Frequency Data value for predicting biomass yields n!ess annual changes in recruitment can be pre­ Total mortality rates can be estimated dicted, at least on a relative basis. One directly from the decline in the relative source of information which can be used to abundance of successive age classes in size estimate the relative size of pre-recruit year frequency data available from samples of classes is CPUE data obtained from commercial catches or landings if it can be research survey cruises which sample juve­ assumed that the age structure of the popu- nile fish before they are recruited to the lation is stable. Stability implies that the fishery. Similarly, size specific or age spe­ relative abundance of successive age cific CPUE data collected from commercial classes at the same point in time is the same landings for a fishery whicn routinely har­ as the relative abundance of a single age vests younger fish can sometimes be used to class if it were followed through time. To predict recruitment to a fishery which har­ estimate Z during some time interval t, the vests older fish. absolute or relative ages of successive size groups must be known. Relative abundance 5.2.2.2(d) Age (Size) Parameters is estimated as catch per unit effort (CPUE). The slope of the natural logarithm of CPUE An estimate of the size at first capture (1c) versus age gives an estimate ofannual total can be obtained directly from samples of mortality when t equals one year for all age commercial catches and can be converted to groups represented in the data. The same age (t,.) if the growth history of the exploited analysis can be performed for the relative species is known. Estimates of size (age) at decline in abundance of a single age group first capture will correspond to size (age) at (year class) through time. In either case, it is recruitment (lr, t,) as long as there is no important to only use age groups which delay between the time that fish first have fully recruited to the fishery and are become available for harvest and when they completely vulnerable to capture. Vulnera- are actually harvested. Delays occur when bility to capture depends on the gear used the fishing gear is size selective (i.e. w1 .i:t and its size selective properties. only harvests fish within a specific size range). When t, does not equal t, estimates If estimates ofthe growth parameters K and of t, (or 1,) must be based on some knowledge L . are available from other sources, total of migratory behavior, i.e. the age (size) at mortality can also be estimated from the which recruits can be expected to arrive on equation derived by Beverton and Holt the fishing grou,,ids. (1956) K(L-) 5.3 SOCIOCULTURAL DATA COL- Z = -C- LECTION METHODS when Ic equals the mean length at first cap- 5.3.1 Introduction ture and 1:-Ithe mean length in the catch for lengths greater than lc. The parameter 1c is Before any data collection scheme is imple­ best defined as the length of 50% retention mented, the investigator should ask himself for gears such as trewli or gill nets where "why are we collecting these data?" It is mesh size is important a,"d selection occurs especially important to ask this question over a range of sizes. when one begins to construct the interview form fcr a sample survey (Casley and Lury formu­ 5.2.2.2(b3) Changes in Total Mortality 1981), but it is also important when and Effort lating questions to be posed to key infor­ mants. In preceding sections, justifications If a series of paired Z and effort estimates pe- were provided for most of the sociocultural are available for a series of equal time ' that of a linear regression c data needs, but it must be remembered ods, the Y-intercept estimate of M. some were situationally dependent. For versus effort provides an data on pro- Forvers efforlpris an esrtmaye o Mexample,y one need not obtain are pro- For this analysis, Z and effort may vary posed innovations if no changes location or even by ageoffish jected. Furthermore, data concerning with time, by or exploited population (see Beverton attitudes towards changes in income in the if these and Holt, 1959). This method assumes that occupation need not be collected time, location, or age remain changes are not expected to be part of the fish at any Obtaining equally vulnerable to capture, fishery development process. this attitudinal data is per­ Recruitment and analyzing 5.2.2(c) haps the most time consuming aspect of any "why are we Yield per recruit assessments are of little survey and thus the question

91 collecting these data" should be posed sev- munity (4.2.4.1). Preliminary assessments eral times when collecting attitudinal data. of attitudes towards risk, change, and investment (4.2.4.2) as well as perceptions of 5.3.2 Examination of Available proposed innovations can also be obtained Records from key informant interviews (0.2.5.4 Usually tiie information that can be found through 4.2.5.8). These preliminary assess­ ininualltte written doumntsindocuments is ltetlimited cto nscensus oud piements, surveys however, of the must attitudes be confirmed and perceptions by sam­ of fishermen. which, if sufficiently detailed, can material and loca­ be used to determine the numbers tions of potential project participants (4.2.2.3), and aspects of the occupational 5.3.4 Data from Sample Surveys The investigator must be structure (4.2.2.6). simple. easy to ask warned that in many cases census material There are a number of in such a case, questions which can provide quantitative does not include occupation; by are available they can estimates to support statements made if licensing records can be to estimate numbers of i'ihcrmen. key informants. These questions be used included in a sample survey interview 5.3.%Interviews With Key Informants which is administered inexpensively during a short period of time. Surveys provide gen­ A great deal of the sociocultural data dis- eral background information as weji as cussed in the preceding chapter can be data which can be used, fir example, to obtained from we'l selected key informants. determine the social structure of the occupa. For example, key informants can provide tion. Fishermen can be asked if they are many of the necessary data such as group presently -employed (4.2.2.7), when and for identification (4.2.2.1), degree of intergroup how long they go out fishing (0.2.2.9), the tension (4.2.2.2), locations of potential pro- kinds of gear and vessels they Wi- and -Nil ject participants and estimates concern,*ng owns the equipment (4.2.3.1), v-hether,,r not numbers (4.2.2.3) which should lie verified they can realistically expect to own a %,.ssel by a review of written records or by conduct. themselves some day (4.2.3.4), and what ing some sort ofcensus. Key informants can their position is in the crew (4.2.3.6), concerning tradi­ also provide information sec­ communication channels (4.2.2.4), Much of the attitudinal data listed in tional and level of community development (4.2.2.5), tions 4.2.4 and 4.2.5 (Innovation Data structure of the target com- Occupational Preference and Training occupational of sam­ munity (4.2.2.6), temporal distribution of Data) must be obtained with the use fishing effort (4.2.2.9), and local knowledge pie surveys. This methodology is essential about fishing and fish (4.2.2.10). because we need to know which individuals or groups will be more likely to accept a Much of the information concerning the proposed innovation. Such conclusions ai P social structure of the occupation of fishing based on the statistical properties of the can also be obtained from key informant sociocultural variables which are !,xam­ interviews. General information concern- ined. For example, once we have determined ing fishing gear types and ownership pat- the history of innovation in the community terns (4.2.3.1) can be obtained from (4.2.4.1) through the use of key informants, knowledgeable local fishermen. Likewise, a widely accepted innovation can be identi­ key informants selected from the local fish- fled, and individual fishermen can be asked ing population can be used to provide data when they began to use this innovation, concerning crew size and social composi- how long they have continued to use it and tion (4.2.3.2), criteria for crew selectien whether or not they still use it. These data, (4.2.3.3), degree of occupational mobility in combination with individual data such (4.2.3.4), patterns of interaction betwf en as age, years of formal education, degree of owner and crew (4.2.3.5), degree of on and literacy and exposure to mass media, isee off vessel specialization (4.2.3.6), relation- list under 4.2.5.9) can be used to determine ship of fishing groups with other groups individual attributes assotiated with inno. (4.2.3.7), and the local distribution of wealth vativeness in the community (4.2.5.9). and power (4.2.3.8). Attitudes towards risk, change, and invest. Finally, key informants also can provide ment can also Ie determined througi the data concerning innovations. Local fishing use of sample surveys. Questions such as equipment suppliers and older fishermen those listed under ,1.2.4.2 could form part of can be used to supply data concerning a a survey questionnaire. The distribution of history of innovative behavior in the com- perceptions of the attributes of a proposed

92 innovation among individuals or groups ble since it is extremely difficult to antici­ must also be determined with the use of a pate answers to these types of inquiries. survey (4.2.5.4 through 4.2.5.8). Individuals Closed ended questions can sometimes elicit could be provided with a description of a responses which are biased in the direction proposed change and asked to comment on of the investigator's expectations - they its complexity, compatibility, the advan- suggest an easy response for the busy tage it provides over the old techniques, fisherman who is anxious to get rid of the whether or not they feel they could actually interviewer. experiment with it, and if they feel that they can adequately assess its potential by Responses to most open ended questions observing proposed demonstration can be categorized into ten or fewer catego­ techniques. ries with one residual category for idiosyn­ cratic responses. For example, in a study of Finally, survey data is also necessary to attitudes towards the occupation of fishing, obtain accurate indications of occupation Pollnac and Ruiz-Stout (1977) asked the preferences (4.2.5.1 and 4.2.5.2) as well as question "what do you like about fishing in attitudes towards income changes (4.2.5.:3). comparison with other occupation3?" Several techniques can be used to obtain Responses were categorized into a total of these data. First of all, potential fishermen seven categories, including a "don't kn N" can be requested to compare fishing with and an "other" category. The open ended realistic alternative occupations. For cxam- responses categorized as "sport-pleasurc," pIe, the following questions could be posed: "monetary reward," and "independence" (a) Would you prefer to be a farmer (for accounted for 83 percent of the responses example) or a fisherman? Why? (b) Would while )rly nine percent of the responses had you like your son to become a fisherman? to be classified in the catch-all "other" cate­ Why or why not? (c) Would you become a gory. Evidence that analysis of open ended fisherman if you had the opportunity? Why questions can be easy is found elsewhere in or why not? Analysis of responses to direct the cited work (Pollnac and Ruiz-Stout, questions such as these can be quite 1977) and other papers by the same authors. revealing. A more sophisticated and time consuming 5.4 ECONOMIC DATA COLLECTION technique would be to comipare fishing with other realistic alternative occupations by 5.4.1 Introduction examining a series of relevant dimensions which represent occupational likes and dis- This eection is not meant to be a comprehen­ likes. Dimensions related to job preference sive treatment of either applied fishery sta­ must be culturally relevant, so they should tistics nor of sampling theory. For a be determined through preliminary inter- thorough dincussion of the former the views. Indivi.&'als could be requested to reader is referred to Bazigos (1974) and for a compare different occupations and tell what comprehensive set of lecture rotes on the they like and dislike about them. The output latter to Chakraborty and Wheeland (1979). of these interviews could be used to con- The discussion which follows will examine struct a list of relevant dimensions such a some of the causes of variation in observed income, status, "goodness," intelligence, values of the many variables discussed in family life, or least preferred to most pre- Chapter IV and, based on this, suggest how ferrPd occupation. A sample of individuals to aggregate the collected data. This will be could then be presented with a ladder dia- followed by a very general discussion of gram and told that thc bottom rung of the sampling in order to highlight some advan­ ladder represents the lowest possible tages of using sales receipts as both a cen­ income and the top rung the highest. They sus of fishing activity and a sampling would then be requested to place each of the frame. occupations on the appropriate rung. This be done for each of the dimensions The advantages of having a census of fish­ would all resulting in a comparison of occupations ing activity available are apparent to along all dimensions. concerned with fishery statistics. However, due to the problems associated with con­ Some people (Casley and Lury 1981, for ducting such a census, little has been wrnt­ example) object to the use of open ended ten about it. It is the authors' belief that questions such as these in surveys because sales receipts can Le used to provide infor­ they are difficult to analyse. On the other mation on fishing activity in much the same hand, open ended questions can be prefera- way as a census.

93 5.4.2 Kinds ofVariables and Sources of some are the result of economic and techni­ Variation cal decisions by fishermen, and others are the results of natural (biological and envir­ Section 4.3 lists a large number of variables onmental) phenomene. useful in many different kinds of economic and bioeconomic analyses. The variables The nature of the variable itself determines were put into two groups: those relating to a to a great extent the number of factors day's fishing by a single firm (called a unit which affect the values it assumes. For of fishing activity) which are best measured example, observations of a biological (or at the end of a fishing trip, and those which related) variable such as weight (or value) of are not related to fishing activity and are a particular species caught on a given day best measured in a community setting. One result from economic factors (the boat/gear use of this data is in estimating annual combination in use), physical factors (how amounts of the quantitative variahlep that frequently the net was hauled), and biologi­ describe the operation of the fishery. We will cal factors (e.g., the dispersion ofthe species continue to use this as a framework in dis- in the area fished). Factors which produce cussing the data because all other uses of variability can operate on a daily, monthly, basis. Weather and pollution the data will be served if one can obtain or annual additional sources of properly grouped estimates of could also be added as unbiased, It ia clear that variations in annual amounts of these variables. Inter- variation. view data will be discussed first. observations of biological (and rclate,1)var­ large number of The total annual amount of any given quan- iables result from a very titative, activity-related variable, such as factors. the amount or value of ice used, the weight or the value of a particular species or the In contrast, non-catch-related economic cost of fuel, can theoretically, either be variables, such as the amount of ice used, directly or estimated. The total and physical variables, such as the time determined sour­ annual amount of variable X is equal to actually fished, are affected by fewer ces of variation. They are, for example, Ex or XoN. exempt from the short run, direct impact of The first case, which is a sum of all of the many biological factors.t The values they theo a assume, of course, result from the various values,Thlersse, is onlyiswhich possiblepssu anly in a oaltheoretical short-run and long-run decisions the fisher­ Also, these economic and physi­ since all values will, in reality, never men make. sense decisions are influenced by how the be known. The second case is an estimate of cal reliability of fishermen evaluate many thebf biological the total annual amount. The factors. is determined by the bias of this estimate mean, the estimate of the true population i.e., R , and the accuracy of the count or estimate of N, i.e., either N or .-"The 5.4.3 Sampling and Aggregating degree of statistical confidence one can have in the annual estimate is related to the A sampling scheme designed to take into variance of the estimator R . This, in t-rn, account some of these sources of variation is related to variation in the observations on can either reduce the total number of obser­ X. It will be seen below that if N is estimated vations needed or, for the same number of as well, then the confidence limits on the observations increase the degree of confi­ estinated total annual amounts are dence in R , the sample mean of the varia­ affected by the variance of this estimator, ble X. An even more important reason for the observations as well. The causes of variance in obser- grouping (or stratifying) vations on X will be discussed first. according to the vaiicus sources of varia­ tion is that the groups or strata themselves The variable X is either a biological, eco- are of interest. In particular, many of the nomic, sociocultural or physical quantita- groupings of the data will represent the tive variable. A very basic question is why results of decisions made by groups of two or more daily observations of X are dif- fishermen who live in similar conditions, ferent.'" There are, of course, many reasons: are organized similarly, use the same boat

By daily observations we mean observations made between days on the same boat or between boats on the same or different days.

I They may be subject to some sources ofvariation from which biological variables are exempt. For example, unloading fees may differ over landing sites.

94 and gear types, and will be the target of ing on location. For example, landing sites may be a strong indicator of area fished, departmental actione."1 home port, gear type, etc. variatio,. only a small Of all the sources of number are important in producing large The number of strata in which the data are variation in several kinds of variables. Still aggregated and stored should be at least as fewer are also easily observable. Those great as the number of strata used for sam­ which do satisfy the first condition are vari- pling.' There are several important sources ables describing qualities of the unit of of variation which are not easily observable activity such as home port of boat, location and cannot be used to stratify the samples fished, gear type, boat class, mesh size, but which can be used to distinguish stored method of propulsion, size of crew, etc. data. One of the most important of these is Many of these are referred to as identifiers the area fished.'" Interviews will produce Each of these qualitative variables data on areas fished but for any particular above. by affects the range of the observed values of a area this information will be obtained but overlapping set of quantitative chance. The probability of collecting data different as col­ variables. Some of these qualities are not by fishing grounds can be increased easily observable, they require much closer lected data are examined for correlations examination or responses to questions between areas fished and different landing about them. Many of them have to do with sites or other factors. aspects of nominal fishing effort. Ideally, data should be aggregated in a way The first basis for stratifying the samples to which: 1)allows for ease of generating sum­ be taken (besides date and location) should mary material; 2) is sufficiently disaggre­ be one which is a reasonable substitute for gated so that many sources of variation can (is correlated with) as many of the sources of be examined; and 3) permits newer, more variation as possible. For common bioeco- refined data to be added at a later date (for nomic sampling purposes Chis factor should example, a species breakdown of the catch, be one which affects the values of both bio- market prices for labor services, adjust­ logical and economic variables. In many ments to crew remuneration, etc.). small scale fisheries this will be the boat­ /g,!ai, combination used. Boats may be 5.4.4 Variation in Numbers of Strati­ classified by type (e.g., canoes, sailing vts- fled Units of Activity : further sels, motorized vessels) and perhapF it sailing vessels over 25 feet or The problem of counting N or estimating by size (e.g., N is the total motorized launches with sleeping quarters, ('R), may be considerable. (days of fishing) etc.). More detailed gear distinctions may be number of units of ac&-vity In Costa Rice the by everyone who fishes over the entire year. more difficult to observe. the same combir -tion was a good indica- Counting or estimating N within boat/gear is rore difficult. tor of days fished per trip, range (areas) of strata used for sampling crew size, the general mix Assuming that one dimension of stratifica­ fishing activity, sampling caught and, in some instances, the tion, say gear type, is in use for of fish gear types, the home port and likely landing site. and that there are three annual total for variable X is now estimated How the data are aggregated (i.e., stored by J=3 and summarized in strata) will reflect, first X = r. N. of all, the sampling strata. Data should be j=l grouped for a given time period (e.g., a month) because of the time dependency of Nj is the number of days of fishing over the many of the sources of variation. It is also year by gear type #1 (j=1), #2 6=2) and #3 worthwhile storing data by location of sam- 6=3). The total number of units of activity ple because of the correlations that exist for the entire fleet for the year, N, is equal to between some important variables depend- Ni plus N2 plus Ni.

analyses attempt to capture. 'It is exactly these decisions and their aggregate impact on the resource that bioeconomic then the raw data can be stored by identifiers, I facilities and programs for data manipulation are available, If computer can be stored by necessary aggregations can be generated at will. If this capacity is not available, summaries and any guesses" as to the way of a filing or cataloging system. The raw data should be kept on hand as well since our "ducated appropriate summaries may not always be correct.

area, hook size and number, etc. ' Several more of these are aspects of fishing power such as mesh size, net

95 There will be, of course, two additional sub- other estimates are three or more times scripts reflecting the Iccation i! ) and the removed from the true N. date (t) of the opportunity to sample. Hence, if we assume three linding sites, In some fisheries, because of the logistics involved and the decentralized location of J3 T=365 L3 landing sites, the only reasonable data col- N = E E E Njtt, lection method relies on observations which i=1 t=1 1=21 are not taken in post-trip interviews. Conse­ of a variable such as Xquently, much of the detailed amounts of and the annualaot total afish, revenue and costs revealed in a post­ iss estimatedeth by trip interview setting are missed. J=3 T=365 L=3 In spite of these drawbacks, many of these X = E E E Rjt..Njtt methods are, in fact, fairly sophisticated in i1 t=1 2i4= dealing with many seasonal and locational variations in the amount of fishing activity. There are many reasons why the number Correlations between time and/or locations and composition of units of activity (Nj) and several other factors such as boat/gear vary from day to day, month to month, and combinations can be and are exploited, and landing site to landing site. Therefore, there many of the infrequently observed details are even more reasons why jtX.Njtxvar- can be estimated to yield accurate results ies. As with X, these variations have daily, (see Bazigos, 1974; Banerji, 1974). The esti­ seasonal and annual aspects which are par- mates of total annual amounts are based on ticularly important in small scale fisheries, estimates of both the variables and the For example, the amount of fishing activity amount of fishing activity, however. The taking place on a given day is influenced by use of sales receipts can solve some of these the weather and the seaworthiness of small estimation problems and provide valuable vessels. Seasonal influences include migra- information for reducing casts associated tions of fishermen and seasonality of agri- with increased accuracy. cultural work. In the long run there is entry into and exit from the fishery as surround- 5.4.5 Sampling and Aggregating with ing economic conditions change. In general, Sales Receipts the list ofsources of variation in the number and composition of units of activity (here The advantages of being able to use sales differentiated by gear type) per unit time is receipts, either simple or detailed, ftor all almost as long as the one describing sources or even some landing sites, for all or some of of variation in variab!e X.There are several the year, and on an indefinite or periodic ways to estimate Njtl. For example, people basis (alternate or every fifth year, for could be placed at a few sites for the dura- example) will become evident from the fol­ tion of the year or at all landing sites for a lowing discussion. Statisticians will discern short period of time. They would first keep even more advantages than are described track of all landings and second try to here. The purpose of this section is to record the number of days of fishing. More encourage consideration of the benefits and reasonably, they would estimate the costs of such a scheme. AE long as the major­ number of days of fishing per-trip or land- ity of the catch is landed within a reasona­ ing. The number of estimated days of fish- bly centralized area and is sold (rather than ing per landing could then be used along bartered or consumed directly), the econom­ with any other information (about relative ics of supplying incentives to primary buy­ number of landings at other sites or other ers to use sales receipts as opposed to times of the year) to extrapolate over all supplying incentives to fishermen to pro­ sites or all months respectively. The result- vide data, should be obvious.' 7 ing estimate of N, however, is at least twice removed from the actual count. Other fre- An ideal simple sales receipt might contain quently used sampling techniques focus on data such as location of sale, value of the boats or landing sites in trying to estimate amount sold, date, a commercial classifica­ annual amounts. Some of these techniques tion of the amount sold (i.e., prices and require that the number of fishing days per weights), a receipt number, name of the boat or per landing site per time period be buyer and name of the seller. This simple estimated as an intermediate step. Still receipt supplies us with an estimate of Njt)

These two options are clearly not mutually exclusive.

96 i.e., an estimate of the number of units of The first advantage is clear; an exact enu­ activity by site by month for the purposes of meration is preferable to an estimation, and calculating monthly totals. We are still the cost ofthis enumeration is smallrelative required to estimate the landings by gear to manpower costs. The second advantage type and the number of days fished per land- is almost the same as that provided by sim­ ing. These can be estimated using the inter- pie receipts, except that the desired number view data if the interviews are not biased in ofinterviews for the next time period can be favor of particular gear types. determined by more strata - most impor­ tantly by boat/gear combination. The abil­ purposes, the simple receipt ity to use'this advantage can be determined For sampling between landing landings by site by day of the week by examining the relation supplies the various strata. week of the month. This can be used to time and location in and will supply t'le data for estimate the amount of fishing activity in Detailed receipts of land- this analysis. These relationships almost the next period and the frequency and sites. For purposes of pro- certainly exist in the artisanal fishery, ings at different of them by sampling, the simple receipt it is possible to take advantage portional procedures - even allows one to determine how many post-trip adjusting the sampling breakdowns should be to the point oftarget sampling. An interest­ interviews or species efficient in addition, it supplies us ing relationship to be examined for attempted. But times and enumeration of an important varia- sampling is between landing with an as thesePxistand are which appears on the interview form: tides. If relations such ble the most inefficient (in amount (and value) sold to the primary not recognized, no need to esti- terms of cost per sample) methods must be buyer. There is, therefore, a amounts. But there are, used to assure an unbiased sample, i.e., mate these annual of personnel course, many reasons to continue to col- completely random placement of time. lect these pieces of data in interviews. The over location and further uses of this variable will be dis­ cussedsupply below.a commercial The sales classification receipt may of alsothe Thenext numberperiod(regardless of samples of totheir be specifictaken in time the total amount sold - helpful information for and location distribution) can be deter­ determining species composition. mined in several ways. The most efficient way uses the number of landings (and not days fished) as the basis. This is because the A detailed receipt might contain, in addi- sampling unit, the landing, may contain tion to the data mentioned above, items data for several days of fishing. Sampling such as timeofsale, boattypeused, method on the basis ofthe number of landings in the of propulsion, obvious gear used, number of previous period will come closer to yielding people fishing, number of day m ,bhetrip, a given percentage of the number of days of boat name or registration number, home fishing than will a scheme based upon fished,hone number of days of fishing in the last period. i.e.,port manyof the boatof the and, same perhaps, variables which The difference is only one of efficiency, kind of however. on the interview form. This appear an exact count of the receipt will provide: 1) various NjtI and the count for several addi- The sample means estimated for any given tional strata in which data might reasona- period are multiplied by the N of that period bly be stored; 2) a greater quantity of data, even though the number of observations at an earlier date, with which to determine that were involved in these calculations the number of samples required for what- were determined by the previous months ever strata; 3) a greater quantity of data, at landings. In other words, the number of an earlier date, with which to establish the interviews for a period is a percentage of an presence or lack of correlations between estimate of the number of landings that both sampling location and time (hour, day period. Various methods can be used to of week, week of month) and the various obtain this estimate. One might simply use strata; 4) the possibility of reducing the the total of last month's landings. Trends in number of interview questions about even the amount of fishing activity from month the most changeable economic variables ­ to month can be accounted for by weighting catch related costs - asked in subsequent the previous period's landings. years of data collection, and 5) the possibil­ the receipt become the primary Better still, one might have available data ity of having of source of data in the future - thereby reduc- on the number of landings for the period ing data collection costs (but increasing the interest from previous years. As time goes data processing costs). on, the accuracy with which we can predict

97 the number of landings for the coming pen- labor costs can be related to the value ofthe od(s) will increase. It is because of the diffi- fish sold, the number of men fishing, and culty in trying to predict the next period's the boat/gear combination. These relation­ activity that we suggested in 4.3 that a com- ships can then be used to estimate values for bination of methods be used to calculate the same variables (e.g., total labor costs) the early for all landings during some period of time maintenance and repair costs in 9 The collection, from data provided on the receipts.' years of data interviews become, in essence, an occasion­ for revalidating the The fourth and fifth advantages are related ally applied instrument to each other. Even innthcaeosmpigintensive the case of sampling relations interv'ewingestablished inand the for firstdiscovering years and f o s aes receipts, the interview data can ines interelins without changes in those relations. The use of sales reveal relations between catch related vari- of a the catch sold to pri- receipts is worthwhile regardless ables and the value of ability to exploit this mary buyers."' For example labor costs fishery department's may be based on several pcssible measures last advantage. inter- of the value of the catch. With enough can be stored separately view data this cost can be related, with vary- The sales receipts ingdegrees of cotfidence, to the amount of but in parallel with the stored interviews, buyer. If a relation- i.e., within the same strata defined for inter­ fish sold to the primary how­ ship exists and is fairly strong, it might be views. Their number will be so great, possible to eliminate questions about labor ever, that a system of summarizing the data costs from the interviews given that we they contain must be established. This have recorded the number of crew. of using a holds for several other variables as well. Of It is suggested that the feasibility if the estimate of the total annual receipt system be evaluated by examining course, compliance, the use of value sold to the primary buyer is biased, the degree of expected to annual catch incentives, the ability to distribute and col­ this bias will be transferred human A simple or detailed sales lect the receipts regularly, and Lhe related costs. resources necessary to receipt system will eliminate the probability and/or computer tabulate, store and interpret completed of such bias. receipts. The logical extension of this elimination of variables is to be able to substantially 5.4.6 Fixed Costs interviews and reduce the dependence on means of for a significant period of time (severai Fixed costs are best measured by receipt data, a regular interviews conducted in fishing communi­ years), rely on sales of fixed costs sampling program to determine the species ties. Total annual estimates catch, and occasional in- can be obtained by a proportional sample of compositior of the number of depth interviews to reveal changing the population of firms. The total firms, or better still, the total number in relations. each stratum, can be derived from a census It is inadvisable, however, to completely or from registration lists, or it can be es'i­ forevo the regular co!ection of economic mated per village, per landing site or per inte''ew data on the basis of relations geographical area. The proper strata for between a given variable and only one other sampling the firms and storing the data are variable (say a percentage relationship principally defined by economic and loca­ even within a grouping of interviews). The tional sources of variation. The sample amount of statistical confidence which can number (a percentage of the estimated total be assigned to this relationship will proba- number) in each strata should reflect bly not be very great. The conditions which defined amounts of capital investment and permit the interruption of economic data common sociocultural characteristics in interviews require a complete year of inter- each grouping. Again, a good indication of view data and computer facilities for per- the amount of capital involved is the boat­ forming multiple regression analyses. /gear combination in use. This is, further. Given these conditions it is possible to relate more, a stratum common with the variable many of the variables compiled from the costs. The timing of these interviews is dis­ interview forms to each other and t9 then cussed in Chapter IV. apply these relations to variables repre­ sented on the detailed receipts. For example,

" After a year's data collection, the frequency with which we need to collect catch.independent cost data is greatly reduced. We should know, for example, the mean quantity of ice used per days fishing by particular boat/gear type (within each stratum). This estimate should have a relatively small variance.

"IThere is a strong possibility that the annual economic quantities estimated from multiple regression relations will be more accurate than those estimated by sample means.

98 Chapter VI

DATA COLLECTION STRATEGIES

6.1 INTRODUCTION lected by one investigator may be useful to another. In either case, substantial savings A major objective of this guide is to demon- can be realized, especially if data are being stratethe importance of integrating biologi- collected from a number of remote locations cal, economic, and sociocultural data in requiring large investments in time and order to generate information which is use- vehicle (or boat) use. Coordination of data ful to fishery administrators in developing collection efforts at this level is not compli­ countries who are responsible for managing cated, but does require thatdata be collected and developing small-scale fisheries. Pre- during the same period of time and that at vious chapters have described information least some of the survey locations be the and data needs a3 well as appropriatedata same. It also requires that investigators collection methods for each discipline. In working in each of the individual disci­ this chapter, we will outline two strategies plines be familiar with each other's data for integrating data collection efforts which collection procedures and schedules. should be applicable in most situations which require multi-disciplinary informa- The second level of integration involves the tion. Before examining these strategies in design and implementation of common more detail let us first discuss the ways in data collection procedures. At this level, which the data collection efforts of different coordinated data collection activities are investigators can be effectively combined to not only useful, they are essential since provide the greatest amount of information investigators in different disciplines are fol­ to policy makers at the least possible cost. lowing the same survey format to obtain a Even greater savings 6.2 INTEGRATING DATA COLLEC- common data base.' are realized than at the first level since each TION EFFOETS discipline can make use of the same inter­ Data collection efforts can be integrated at viewers, sampling schedule and sampling two levels. At the first level, the data collec- locations and thus avoid a great deal of tion activities which are planned and car- duplicated effort. Collaboration at this level ried out independently by biologists, requires that research and data needs be economists, and anthropologists/sociolo- clearly defined before data collection be­ gists are coordinated in order to save time. gins. In addition, considerable time must be In its simplest expression, an interviewer devoted to the design of procedures which who is responsible for collecting, say, catch will produce as much multi-disciplinary and effort data for resource assessment pur- data as possible, but which are not so poses may be asked to also obtain price unwieldy that they are difficult to adminis­ information or compile lists of fishermen for ter. For example, it is particularly impor­ Interviewers tant to avoid using lengthy survey later sociocultural interviews. to conducting economic or sociocultural sur- questionnaires which take a long time veys may also be able to provide valuable complete. In a statistical sense, the success­ information for the biologist. On the other ful design ofjoint data collection procedures hand, some data which are routinely col- requires the identification of common sam-

SIt is unrealistic to assume that all the data required for even the most compatible purposes (e.g. stock assessment and variable cost and earnings analysis) can ever be collected using common data collection strategies; there will'always be some additional information which must be obtained independently.

99 pling units and a sampling scheme which 6.3.1 Strategy Based on Static Frame produces estimates that are unbiased and Surveys reasonably precise. This is not always a simple matter when, for example, a biolo- In the approach described by Bazigos and gist estimates mean catch per species and Banerji, total catch (and catch by species) is an economist estimates the mean cost of a estimated from a "catch assessment sur­ day's fishing, even though both are equally vey" which is designed on the basis of a interented in knowing how much fishing census of sampling units (usually boats effort is associated with a given catch or a and/or landing sites). This census or frame given cost of operation. survey represents the number of sampling units which are counted at a particular In seeking to attain a greater degree ofcoop- point in time and must be repeated periodi­ eration between researchers who are study- cally as the number of units in the fishery or ing the small-scale fishery system, it is their distribution changes over time. Total obviously preferable to reach the second catch can be estimated from the mean catch level of collaboration. Data collection proce- per boat or per landing as long as an esti­ dures which are described in this chapter mate of the total number of boats or land­ are designed to achieve that objective. This ings is available. Estimates of total fishing ia, however, an ideal objective which isn't effort (hours fished, for example) can be always attainable because 1)a strong com- obtained following the same procedure. mitment to cooperative research does not Furthermore, the catch assessment survey aiways exist, even when multi-disciplinary can easily be extended to generate economic studies are undertaken, or 2) given the diver- cost and earnings information. Gains in sity of sampling objectives, each with itr precision are possible if the information own associated statistical variability, data available from the frame survey is detailed collection methods are not always compati- enough to allow for the stratification of the bolectinD es thes ble. Despite these problems,prbems, wewt liebelieve thattat- phicalsample region. survey Althoughby gear type it is notand/or our purposegeogra­ multi-disciplinarylection are needed andapproaches should beto promoted.data col- tophicescre describe thisthouhprocedurei innt detail,u p we doo lectonanre shuldeede e pomoed. offer the following brief summary of how it Of the two strategies discussed in this chap- ofe te follow n ef ao f wit ter, one ir. particular is designed for achiev- could be implemented along with an ing closer cooperation between the three example: disciplinev which are represented in this guide. It is up to the reader to consider how Step 1: Define the sampling unit (e.g. indi­ other issues mentioned in Chapter I might vidual boats at individual landing sites, be included in an even more comprehensive preferably categorized according to gear and unified data collection scheme. type); Step 2: Conduct a census survey of sampling 6.3 TWO ALTERNATIVE DATA COL- units within a defined geographical region LECTION STRATEGIES (or regions); There are two recommended approaches to collecting catch, effort, revenue, and cost Step 3: Randomly select some proportion of data which apply to small-scale fisheries, all units for sampling during a defined Neither one is betio- than the other, but one period of time (e.g. a month), choosing an may be preferable over the other in a partic- appropriate sample size (e.g. 10-20% of all ular situation. Although the two boats) acording tothe desiredlevel ofpreci­ approaches are similar in many respects, it sion and the amount of manpower, etc. is important to understand the differences available to conduct the survey; between them so that the appropriate stra­ tegy can be selected when necessary. The Step 4: Either station someone at each first approach has been described in detail selected landing site to obtain catch, effort, by Bazigos (1974) and Banerji (1974) and is variable cost, and earnings data from ran­ widely followed in a number of countries; domly designated sampling units or devise the second approach is not so common, but a schedule for visiting individual sampling may be more applicable in certain circum- sites and collecting the same data randomly stances which will be described below. It is in each site; our contention that the second approach, which is described in more detail in this Step 5: From mean catch, effort, revenue, guide, is more amenable to multi- and variablecostsperboatperlanding, esti­ disciplinary data collection needs. mate total catch, etc. by extrapolation.

100 As an example of how this data collection in terms of the production activity of a firm procedure could be applied in a real situa- (e.g. days fished by boats using a particular tion, consider the case in which three land- gear). Also, since total catch, effort, and ing sites are randomly selected within a revenue data are recorded by dealers, certain geographic sector every week. The government fisheries agents have more sampling unit is defined as a landing by an time to devote to intensive landings individual boat in a given site and the frame surveys. survey consists of an enumeration of all sites within that sector and the This approach has been described in pre­ landing basis of number of boats actively involved in small- vious chapters of this guide on the at each site. The frame survey its utility in multi-disciplinary evaluations scale fishing to the would have to be up-dated periodically by of small-scale fisheries; itis preferable the census in a portion of all the methodology described previously (section repeating fishery is landing sites in the sector. Observations are 6.3.1) in situations where the made at each selected site for two consecu- fairly small and restricted to a well-defined every week; the interviewer counts !ographic area, when most of the catch is tive days exist­ number of boats landing during sold to dealers, and when there is an the total sales those two days and determines the landed ing or potential practice of recording catch of a subsample of all boat-landings information. landing, selected systematically (each and every fourth landing, The estimation of total catch, effort, every other landing, and the use of random start. The estimated revenue from sales receipts etc.) with a surveys to generate monthly catch at each selected landing site intensive landings weighted aver- additional information such as the species is obtained by multiplying a effort three selected sites times composition of the catch, improved age catch for the estimates form of landings at all sites as estimates, and variable cost the total number more comprehen­ from the frame survey, a major part of a larger, determined sive multi-disciplinary data collection stra­ suiveys based on a static frame sur- tegy which has been described in this guide Sample the of boats and landing sites are probably and which will be summarized in vey strategy the only alternative in situations where remainder of this chapter. This there is a large number of fishermen or consists of five basic steps. boats in the fishery and/or when fish -is landed in a large number of remote sites Step 1: Compile General Background along the coast. This approach is also pre- Information ferable in situations where buying and sel­ ling arrangements between fishermen and In situations where little or nothing is dealers are informal and where there is no known about the fishery, the collection of tradition of recording catch and revenue general backgound information ideally pre­ data when fish are landed and/or no hope of cedes the collection of more specific data. promoting such a practice (see 5.4). This information simply describes the fishery system and how it f'inctions in very 6.3.2 Strategy Based on Continuous general terms. It is based on the observa­ Frame Surveys tions and experiences of people who are with the fishery ane is fairly easy involves the use of sales familiar This approach fish is to obtain. receipts ­ which are filled out when landed and sold to primary dealers - to An extensive list of descriptive information estimate total catch, effort, and revenue and needs is presented in Appendix A. This list as a frame survey of fishing activity upon is by no means complete and does not which to base a random landings survey include all the types of information which which generates more detailed catch and might be required in any given situation, effort data as well as data on the costs of but it does indicate the extent and the var­ harvesting. The major advantages of this iety of the information which is included in approach are that 1) total catch, effort, and this category. The information needs pres­ revenue are recorded directly as fish are ented in this list relate only to resources and landed and do not have to be estimated from their habitats, harvesting activities, and a landings survey, 2) the frame survey is the contexts in which the fishery exists. continuous in time, geographically com­ plete and does not require periodic up- General background information can be dating, and 3) the sampling unit for used to identify the important activities and more subsequent landings surveys can be defined elements of the fishery which require

101 intensive data collection efforts. It also pro- In any event, a distinction should be made vides the basis for forming perceptions and between recording data - a job which is hypotheses about the fishery and can there- done by the dealers - and retrieving data. foi a be used to design more intensive data Data retrieval requires either that the deal­ collection methods. Failure to collect ers send copies of their records to govern­ enough background information before ment officials or that government agents implementing a data collection program collect copies of the receipts from the deal­ can result in poor sampling designs, unreli- ers. If there is no existing system of record able data, and the loss of considerable time keeping which can be relied upon to provide and effort. Furthermore, the collection of the necessary data then a small group of background information provides an oppor- dealers should be randomly selected and tunity for fisherie3 staff to get out of the asked to test a new procedure for recording office and into the community where they transaction data. This trial period provides will gain valuable experience which will an opportunity to determine the feasibility facilitate subsequent data collection efforts. of introducing a new receipt format Background informaticn can be obtained throughout the entire fishery. If the use ofa from written records, by simply observing receipt system c'oes not prove to be feasible, what goes on when fish are caught, landed, then this strat4:gy should be abandoned in and sold and by interviewing people who favor of the strategy outlined previously are familiar with the fishery. Methods for (see section 6.?.I). collecting backgrounid information are sum- The use of a ,otandardized receipt which marized in section 5.1, and additional infor- includes all the necessary information and mation can be fbund in the introduction to the same units of measurement (see Appen­ Appendix A. The presence of a trained dix B) is essential and can be encouraged if anthropologist is crucial at this stage in the government assumes the responsibility for data collection process since he is needed to printing, distributing, and collecting train interviewers, identify key informants, receipts. Approaches which rely on volun­ and supervise the collection of sociocultural tary cooperation between government background information, agents, dealers, and fishermen are prefera­ Step 2: Institute System to Estimate Total ble to approaches which require Catch, Effort, and Revenue From Sales compiance. Receipts A major problem which must be confronted The estimation of total catch, effort, and when recei ,4s are relied upon to provide esti­ is is a crucial step mates of total catch, effort, and revenuerecorded. revenue from sales receipts must that some transactions are not data collection process. Efforts in the case, some estimate must be made early to institute a system of col- When this 's the transaction data. The be made o the percentage of total landings lecting and compiling (or reported). Other feasibility of using receipts to generate which are not recorded If problems which have already been dis­ these data should be carefully evaluated. V already using receipts to cussed in greater detail in Chapter most dealers are correct landings data record the quantities and value of fish include 1)the need to from the fishermen, it in order to obtain catch data, and 2) the fact which they purchase be very crudely esti­ may be a fairly simple matter 0 encourage that fishing effort may the same procedures. mated. Unless fishing effort is recorded on other dealers to follow of days fished (or Existing procedures should be examined for the receipts as the number In many cases, t&e hours) per boat per landing, the only possi­ possible improvements. be the number of receipt forms which are used may vary so ble estimate of effort will or be so crude that transactions, which, in most cases, can be much between dealers equal the they will have to be replaced altogether. assumed to approximately retrieval may result from 1) number of fishing trips made by individual Improved data of "days fished" on recording more data on each receipt, 2) boats. The recording standardiz- each receipt is crucial if information from recording more reliable data, 3) a random form which is used so all the receipts is to be used to design ing the receipt for obtaining economic dealers record the same types of data, or 4) landings survey t s from more dealers. information based on the production activ­ collectingdealers morerec receipts fity of each firm type (see section 5.4).2

2A day's fishing is most easily defined in terms of a 24-hour period away from port or any fraction thereof and includes any activity during that period which incurs a cost (e.g. travelling to and from port, catching bait, buying ice or fuel). A boat, for example, which leaves port at 3 PM the first day and returns before 3 PM the second day was "fishing" one day even though it may have been underway, and not actually fishing, for fou;" hours. If it rurns at 8PM, it would count as two days fishing. The system used must be simple enough that it will be readily understood by the dealers who are filling out the receipts.

102 Step 3. Other Sources of Resource Assess- Step 4: Estimate Catch by Species, Effort ment Data and Variable Costs Fro,,, a Landings Survey If catch and effort estimates are not availa­ the maxi- Once a system for compiling total catch, ble or can not be used to predict sales receipts of exploited stocks, effort, and revenue data from mum sustainable yields landings for collecting other has been implemented, a detailed alternative procedures to determine the assessment data should be survey is required in order types of resource oflandings within indi­ at this stage. If assessment data species composition considered price categories, to improve effort of any kind are available, plans should vidual for estimates for individual gear types, and to probably be made to initiate a system differ­ as well as obtain detailed variable cost data for collecting catch and effort data to deter­ since it will take ent firm types and to use this data vital population statisics different cost accumulate enough mine relationships among the at least several years to variables can be an analysis and since there is variables. Totals for most data to permit using sample means and the no guarantee that once enough data are extrapolated reliable number of days fished. Receipts will provide available, the analysis will produce some the principal the ability to actually count totals of estimates of MSY for all of relations even any ofthem. Since variables. Alternatively, multiple exploited species or data can be app­ of the vital statistics needed for established from interview many receipt data (see 5.4). assessment purposes can be generated in a lied to sales year's time and do not require expensive Te obtain unbiased parameter estimates, and sophisticated research techniques, it any landings survey must be randomized to makes sense to proceed with techniques for account for variations in the more impor­ estimating population parameters such as tant factors which affect the inputs and out­ growth rates, total mortality rates, and size puts of production. These factors have been descri]d in detail in Chapter V. In reality, and age at recruitment and first capture at random sampling time that catch and effort data are designing a completely the same scheme is an impossible task since there are being collected from receipts and landings surveys. more variables than can be included in a single sampling scheme and since it may be Some of the parameter estimation tech- logistically impossible to follow a random niques which have been described in this sampling procedure when it is time to guide are 1)size frequency analysis, 2) anal- assign interviewers to particular sampling ysis of growth rings on scales or otoliths, 3) locations and times. Also, it is frequently tag and recapture studies, and 4) explora- difficult to identify which are the most tory fishing surveys (see sections 3.2.4.4 and important variables and to understand 5.2.2.2 for more information). The first two exactly how they interact with each other methods should be considered first since and how they affect production activity. fish can easily be measured and scales and Additional problems arise if different inves­ otoliths collected when landings surveys tigators pursue different objectives and are being conducted for other purposes. In design random landings surveys based on fact, most - if not all - of the supplemen- different variables. Conflicting objectives tary data needed when compiling size meas- and designs can lead to differences in the urements (e.g. fishing location, gear type) units which are sampled, the sampling are routinely collected as part of aily land- procedure which is followed, and the ings survey (see Appendix C). Another good number of samples which are collected. reason for combining size measurement activities with landings surveys is that A good strategy for avoiding these conflicts landings do not have to be randomly is to define a common sampling unit and selected in order to measure fish. Thus a follow a common sampling procedure which randomly designed landings survey can will provide all the important data needed proceed and incorporate measurement by the biologists and the economists in suf­ analysis. efforts as time permits and when there are ficient quantities to permit fish to measure. In addition, since Researchers must also accept the fact that enough population biologists are trained to quickly recognize they can not exactly define the species when performing size which is being sampled, nor can they expect individual or be measurements, their participation as to obtain perfectly random samples of a survey team would facilitate very sure how good their random sampling members and collection of reliable species-specific scheme is. Therefore, they can only try the a min­ catch, effort, and cost data. reduce sampling bias and error to

103 imim by concentrating on those factors interview and how often boats are landing; which can be quantified in the frame these decisions can be made by the survey survey, crew. The important thing is to begin the sampling effort with a randomly selected Following the strategy based on the use of landing. sales receipts as a frame survey, the prim­ ary sampling unit can be defined as a land- Sample size should be determined on the ing by a single firm. In most cases, the firm basis of the most variable factor, i.e. the will be a boat and crew which uses a particu- factor which requires the largest number of lar type of gear. The secondary sampling samples. This is likely to be the landed unit is the aggregation of landings by firms weight (and revenue) per species; the varia­ (boats) in individual landing sites. Landing bility in some costs per day fished data will sites can be selected for sampling on the probably be much smaller. Those data basis of the number of landings within a related to catch, of course, will vary as much given period of time as determined from as the catch does. Thus, sample size should sales receipt data and then a pre-determined be large enough to generate reasonably pre­ number of landings are sampled in each cise catch data by species. landing site during the sampling period. :' A convenient sampling period is a month. The interview form and procedure which Sampling locations can be selected on the will be used should be tested during a brief basis of the distribution of landings in the pre-survey trial period of a week or two: preceding month or during the same month inappropriate or ambiguous questions of the previous year or on some combination should be eliminated or rephrased and the of the latter weeks of the previous month, interview should be shortened or divided etc. (see 5.4 for details). into several parts if it takes longer than half an hour to complete (see section 5.1.5). A Care must be taken to spread the sampling trial period is also important for determin­ effort within each landing site and between ing how random the survey really is and landing sites out over the course of the how feasible it is to follow the prescribed entire sampling period in order to avoid tak- sampling procedure. Poorly designed sur­ ing too many samples during particular veys are more likely in situations where times of day, days of the week, or weeks of important background information is either the month.4 It is possible, given tho nature not available or has been ignored. In all of the information obtained from the cases the "ideal" sampling scheme will receipts, to include the temporal distribu- emerge gradually. tion of landings (or days fished) in the sam- The duration of the sampling effort depends pling schedule. If this is done.Tedrto ftsapngfodpns considerations should be given to using on the variability of the parameters which lunar (28 days) months instead of calendar are being estimated over the course of a months. year, and, to a lesser extent, by their varia­ bility between years. Variability within In the absence of a sampling schedule, a years will be greater, in most cases, than reasonable method for allocating the neces- variability between years. In most situa­ sary number of samples randomly through tions there is no existing information which time is to divide the total sample quota for documents, say, seasonal changes in spe­ each landing site into four equal parts (for cies availability, operating costs incurred each week of the month), begin each week's during fishing, or income. A good strategy, sampling in a given site during a randomly therefore, is to continue to sample landings selected day ofthe week and hour of the day, for an entire year and to use that informa­ and decide ahead of time how frequently to tion to design a system of periodic sampling sample, i.e. each successive landing, every limited to particular times of year during other landing, cvery fifth landing, etc. subsequent years. Thus, a continuous land­ Many of the practical decisions such as how ings survey in the first year should reveal frequently to sample will depend on factors variations between seasons, and periodic such as how long it takes to complete an surveys in subsequent years would reveal

:1Recall that the unit of analysis, the day's fishing by a firm, will be the basis upon which we calculate our sample means within the strata, while the occurrence of landings by firms will be used to determine the number, location, time, and type of samples we will collect. In the Gulf of Nicoya fishery the magnitude and distribution of landings varied little across the weeks in a month, but there were considerable differences in landing frequency and in the type of firm landing according to the day of theweek and hour of the day.

104 variations between years for the same sea- beginning of the data collection process. sons or months. More specific sociocultural data can be used to improve biological and economic data Step 5: Conduct Sociocultural and Fixed collection efforts. In addition to facilitating Cost Surveys. the work of other investigators, inter" ,iews with fishermen may also provide long-term and economic data such as the costs of repairs interpretationAt this point ofpreliminary the data which analysis have been and maintenance of gear and vessels which of are not obtained during short-term landings collected will lead to the design of surveys surveys. and their assets. While socio- fishermen may be con­ cultural and fixed cost surveys ducted independently, the population they seek to sample is the same. Some proportion of the population of fishermen which is representative of the population of fisher­ men in the nature of their fishing enterprise is the target for fixed cost interviews. Gener­ ally, the target sample for the anthropolo­ gist/sociologist will encompass this group. While the time distribution of these inter­ views is not important (except perhaps for maintenance and repair cost considera­ tions), all geographical areas should be covered. Questions about indebtedness and invest­ ments in gear, etc., are ideally asked after some rapport has been established with the respondents. Sociocultural interviews are well suited for establishing this rapport. Well conceived questions about fixed costs, in turn, convey the impression that the interviewer is familiar with fishing enter­ prises. Those competent enough to conduct sociocultural interviews should be able to ask a coherent set of questions about invest­ ments. The reverse does not hold for those trained only for economic data collection. Responses to questions about investments, obviously, reveal something about the cul­ tural context of those investments and of the future orientation of the fishermen and are therefore of value to the anthropologist. The economist can likewise benefit from the anthropologist's data since they relate to the organization and conduct of the fishing business and to its potential for change. Just as the bioeconomic value of catch, effort, revenue and variable cost data i6 increased by sampling the same units of activity, so too the mutual value of socio­ cultural and economic data is enhanced if the data relate to the same individuals. As pointed out in Chapter III, an equally important function of the anthropologist­ /sociologist is to provide insights into aspects of human behavior which will help the biologist and the economist obtain bet­ ter data. Some of this information is descrip­ tive and should be collected in the

105 APPENDIX A Artisanal Fishery Background Information Introduction: The following is a list of background information which should ideally, be at hand, albeit on a very general level, prior to planning a data collection scheme. The information listed simply describes the fisiery system and how it functions in very general terms. As noted above, it is based on the observations and experiences of people who are familiar with the fishery and is fairly easy to obtain. It is the kind of information which can be obtained from written mater'ials and key informant interviews as these techniques are described in Chapter V. As the data coller,; a proceeds, however, the accuracy and level of specificity of information concerning the various elements listed belo. bhould and will increase. For example, the salinity tolerance of various species is listed as background information. Information on this item will probably be extremely general when the research begins. The background information may be as general as noting that the species of interest abandon the estuary after heavy .aias. Further research, at a later date, will -,rovide more specific information concerning salinity tolerance, but it is not essential as a preliminary bit of background information. It should be made clear that much of this type of information can or should be able to be derived from a general meeting of fishery inspectors and other department personnel who are familiar with the fishery. The list of information types can be reviewed and filled in where possible. In areas where information is lacking, inspectors can be assigned five or ten bits of information to be collected from knowledgable key informants in the fishing communities. This could continue until all of the involved personnel are satisfied that they have a general, not necessarily a detailed, knowledge of all the listed information at the beginning of the data collection process.

1. Habitat Element Length, width, draft, superstructure configuration by type, capacity of holds

Physical description of habitats occupied by principal Useable life of boats species Cost of boat-type new and import duties Source.; of primary production Common vessel improvement practices Rates of primary production in different habitats and Range of boat-type resale values (high, Iov',r- moderate) at different times of year Boat building and repair sources of nutrient input to Sources, amounts and seasonality coastal ecosystems Hull materials used and their supply sources Principal coastal oceanic and tidal currents Sources of training for repairs and construction Tidal amplitude and periodicity History of evolution of boat-types used How number of boats has changed in recent past 2.Resource Element Ownership patterns- owner operated (owner-captain), absentee owner, cooperative ownership Number of species exploited (approximate) Identification of all methods of propulsion Most abundant species Number of each (approx.) by location Most valued species Melor types used - horse power, fuel used, and other Length and weight ranges of exploited species characterist s Scientific and common names of exploited species Useable life of motors Ease of correct identification Cost of new motor types and import duties Temperature and salinity tolerances of principal spe- Range of motor resale values cies (approx.) Sources of motor purchases Ecological niches occupied by principal species (feed- Sources of spare parts ing habits, modes of reproduction, migrations, etc.) Sources of repairs species seasons and locations of principal Spawning Sources of training for repairs abundance of principal History of changes in size and species Ownership patterns of motors Common combinations of boat/motor 3.Harvesting Element Numbers of each combination (approx.) Inputs (Fixed) Speeds of boat/motor combination

Identification of all boat-types Average cruising range of each combination Number of each (approx.) by location Physical characteristics of other modes (sails, etc.)

107 Previous Pa eBc Useful life of other modes Average number of lost fishing days for all mainte­ nance per year Cost of other modes new A.erage number of lost fishing days for al repairs per Source of supplies and repairs for other modes yetr Scurce of training for repair and construction Average number of men per boat/gear-types Identification of all gear-types used Use of family relations as crew Numbers of each type used (approx.) by location Sources of training for fishing, navigation skills Physical characteristics of gear-types - length, width, Use of ice in various boat-types mesh size, number of hooks, etc. Various forms of ice used How quantity of ice is determined for a tip CoatCosft of gear-typeieof gear-type new and import duties Price per standard unit of ice (if any) of gear-type resale value Range History of costs of ice repair sources Construction and Number of sources of ice Materials used and their sources (and import duties) Ownership of sources of ice Source of training for construction and repair combinations Histryf ger-tpesFuel f evluton capacity of various bat/motor History of evolution of gear-types Sources of fuel Ownership patterns of gear-types Price per gallon Patterns of gear-type use among different groups of fishermen History of price changes Common combination of gear with boat-type and boat- Amount consumed per average trip /motor, boat/sail, etc., combinations Oil sources Numbers (approx.) of these combinations by location Oil prices and their history Multiple gear-types owned by a single fishermen or Amount consumed per average trip single boat Typical baits per gear-type Use of insurance for boat, motor, gear, crew, other Bait sources Sources of insurance Prices of different baits and history of changes Costs of insurance Standard food coats per average trip Sources of credit for boat, motor, gear Who supplies food for trips Interests rates for equal !nanlife from different sources Average repayment perid Additional trip costs, e.g., unloading fees by location Systems used to divide trip costs such as the cost of Registration f food, ice, fuel, oil, bait, etc. among*owner, owner­ License fees captain, captain, crew members Association (e.g. Cooperative) fees Time of day/night fishing by gear-type Existence of taxes on income, property, other Frequency of trips by gear-type in a week, month Mooring fees Average length of trip by boat/gear-type Other fixed fees, e.g., accounting costs, legal fees Seasonal differences Major fishing areas Process Relation between home port and fishing area

Most frequent kinds of maintenance to the typesAreas byreserved tradition for certain fishermen or certain gear­ hull/motor/gear Time and distances to reach fishing sites Average (typical) cost of maintenance p- year to hull/motor/gear Usual time for fishermen to return to port Timing of major and minor repairs by month Specific methods of using gears Who pays for maintenance and repair Several gears used on a single trip Who supplies the labor for maintenance and repairs to Gears directed at particular species hull/motor/gear Differences in catch among gear-type (species, sizes)

108 Competition for fish with industrial fleets How is the fish that is sold used - sold fresh, frozen, fleet p (canned, meal), exported Supplies of inputs or prices affected by industrial purchases. Other services the primary buyer provides to a Number (approx.) of days fishing lost per month or Are fishermen required by these services to sell year due to bad weather particular buyer Do primary buyers or transporters give preference to Is this seasonal industrial fleet catch or by-catch for major species - falling, Recent hitory of catches increasing, etc. Other Fishermen's perceptions of primary buyers and their

profit margins Fish preparation methods Fisherman's perceptions of reasons for large differen­ Number of primary buyers, locations e in the prices they receive and retail prices Reasons fishermen select a particular primary buyer Existence ofand services performed by cooperatives or organizations Do some fishermen transport the catch ofothers to sale fishermens members and number of active members Commercial classes of fish Number of the cooperatives own Determinants of commercial class What do training for cooperative officers Rules buyers use to determine if a fish imacceptable Sources of employment opportunities for fishermen Fish sales at sea or to other than primary buyer Alternate cultural, to entry into the of underutilized species Barriers, e.g., physical, Existence fishery agricultural Reasons for underutilization Seasonality of fishing and relation to Tax on catch - percentage of fish caught actually taxed seasons Existence of support prices for fish System of laws pertaining to fishery, e.g., licenses, reg­ trip reports, insurance require­ various species/classes istrations, inspections, Current prices of the ments, inspection of landed fish, pollution restrictions, Price changes (approx.) through time season or area closures, industrial fleet interaction, gear use restrictions Large seasonal changes in prices Number oflanding sites- location, clusters, remoteness Competition between artisanal and industrial catch or from buyers by-catch land- Pier construction and locations Relative magnitude of industrial vs. artisanal deeper draft ings of same species Ability of landing sites to handle larger, industrial catches boats Point of entry in delivery system of or by-catch Are mooring and shelter sufficient -roads, communi General quality differences in artisanal and independ- Infrastructure serving landing sites transport., etc. ent catch of same species cations, centralized markets, public of his job Systems used to divide revenues between the owner, The fishermen's perception owner-captain, captain, and crew The fishermen's perception of what others (non­ of his work Differences in systems by home port and boat/gear fishermen) think combination Attitudes (general) among groups offishermen to adop­ boat-types Fish used to pay the captain and crew, other costs tion of different gear-types and by certain groups Revenue for services other than fishing, e.g., towing, Traditional areas for fishing transport, etc. Traditional barriers to extended fishing trips Are the costs of food, fuel, ice etc. paid before the trip or Traditional barriers to more frequent trips deducted from the revenue afterward The role of the family in the fishery enwrprise Who pays for losses, e.g., gear lost during a trip Fish withheld from sales so that it can be smoked. /dried/salted for later sale by season Species saved for curing

109 APPENDIX B Detailed Sales Report

SALES RECEIPT (1)No. (2) Date (3) Company or Buyer's Name

(4) MR Sellers Name

(5) Hour (6) Place of Sale

(7) Name of Boat

(8) O Type 1 0 Type 2 O Type 3 0 Canoe 0 Other

(9) 0 Gillnet 0 Handline 0 Longline 0 Other

(10) Number Fishing (11) Days Fishing

WEIGHT CLASS/SPECIES PRICE VALUE

1st CLASS (12) (13) Species 1 Species 2 Species 3 (14) Other Total (15) (16) 2nd Class Species I Species 2 Species 3 Other Total 3rd CLASS Species I Species 2 Species 3 Total

(17) All Classes (18) This version of the detailed sales receipt contains: (1) (10) The number of people on the boat; the preassigned number of the receipt - all of the color (11) The number of days of the trip; coded copies of the receipt have the same number and (12) Weights of identified predominant species in each the number series given to the primary buyer is on class; record at the fisheries office; (13) Names of identified predominant species in each (2) The date of sale; class; (2) oretheecompan;(14) The total class weight; (3)The name of the buyer or the company; (15) The class price; (4) The name of the ranking fisherman; (16) The value (price x weight) of that class; (5) Hour of the sale; (17)The total weight of the catch sold; (6) Place of sale; (18) The total value of the catch sold; (7)N orfme registration number of the boat; Depending on the degree of cooperation expected, ad&.­ (8) Spaces for indicating the type of boat- size ranges tional items could be added, (location fished, for exam­ could be substituted; pie,) or deleted. Receipts used in Costa Rica measured (9) The apparent gear used, i.e., what the recorder sees; 101/2" by 51/" and were produced in triplicate. 110 APPENDIX C Combined Post-Trip Interview and Catch/Effort Survey

I. GENERAL Date: - Hour: - AM - PM Location of Sale: Frimary Buyer: Name of Interviewer: Name of the boat: _ Type: Kind of Propulsion Who is the owner of the boat? What is the boat's home port? Who is the owner of the motor and/or gear?

II. EFFORT AM___-PM What date and hour did you leave Io go fishing? Date: Hour. - Total number ofdays fished? - How many people worked on the boat? What types of gear did you use? A_ B On the first day of fishing: Where did you fl&;h with gear A? (be specific as possible) How many hours? When? from until Where with B? How many hours? from until from Day 2: Where with A? How many hours? until Where with B? How many hours? from until Where with A? How many hours? from Day 3: until with B? How many hcurs? from Where until from Day 4: Where with A? How many hours? until How many hours? from Where with B? unil How many hours? from Day 5: Where with A? until Where with B? How many hours? from until

If a gill net was used: How was it used? At the surface? At the bottom? Drifting? Anchored? What is the stretched mesh size? How long is it? How high is it? What kind of twine? If a hand line was used: What size of hook? Number of hooks per line? Number of lines? Kind of bait used? If a long line was used: How many long lines? - Total length of long lines? Total number of hooks used? Kind of bait? Bottom? How was it (were they) used? Surface? Mid-water? Other kind of gear: Barrier net: Size? Beach Seine: Size? Other: Physical Dimensions

111 III. CATCH AND REVENUE Sale to Primary Buyer Commercial Class Weight Price Value Rejected from sale Species (kg or Ibs) (kg or lbs) ...... First Class Species 1(Name) Species 2 " Species 3 " Class Total

Second Class Species 1 (Name) Species 2 " Species 3 " Class Total

Third Class Species 1 (Name) Species 2 Species 3 " Class Total

...... Other _ _ Species I (Name) Species 2 Species " Class Total

Total weight sold (all classes) Total value sale Are you selling anyone else's catch? How much of this is yours: Wt. Value Value of other catch 1.Did you sell any part of your catch before coming here? (If no go to #2; if yes, ask the following questions): What species or commercial classes did you sell and what was their weight and price? Species or class Weight Prize Species or class : Weight ; Price Species or class ; Weight ; Price How much money (total) did you get for that sale? Whom did you sell it to? 2. Are you selling now or did you sell earlier any seafood besides fresh fish? (If no, go to #3; if yes, ask the following questions): Dried or smoked fish? What class or species? ; Weight? Total value What other food (e.g. turtles, clams, lobster, etc.) Class or kind? ; Weight? ; Value 3. What is the total amount of fish from this trip not being sold? 4. What do you intend to do with this seafood? (i) Dry or smoke it to sell later? Weight? What species or class? Weight? What species or class? Weight? What species or class? Weight? (ii) Are you going to sell it later today? Species or class Weight Species or class Weight

112 (iii) To give to the crew as part of their pay? Class Weight (iv) To eat in your house? Class Weight (v) To he thrtwn away? Class Weight (vi) To -s..y far things or services? What Class Weight What Class - - Weight (vii) To give away? Class? - Weight? 5. What other income did you receive on this trip (e.g. from cargo, passengers)?

For what? How much money did you receive? 6. What is the system for dividing the revenue of this trip?

IV. COSTS C, -THIS TRIP Costs Independent of Catch Is the captain also the owner of the boat? How many people went on the fishing trip? What costs are shared costs?

Fuel How much fuel was actually used on this trip? Price per gullon or liter? Therefore, the total value of fuel used was Of this amount, the boat's share is The crew's share is Others are responsible for

Oil What amount of oil was actually used? Price per quart or litre? Therefore, total value of oil used is Of this sum, the boat is responsible for The crew for - Others for

Ice What anwount of ice was used (bought)? Price per unit? Total value of ice used was Of this amount, the boat is responsible for The crew for Others for

Food What was the value of the food you ate on the trip? Are food costp shiared?. How?

113 Other Costs What other costs were there for this trip or for unloading and selling the catch? Unloading? Cost Sorting? Cost Cleaning? Cost Other? What? Cost Are there any costs that you still have to pay? What? Are any of these cost shared? How? Are any of these costs related to the amount or value .f the catch? How? What is the system for dividing costs of a trip on this boat? How many times have you landed fish in the last 7 days? In the last 30 days?

Maintenance & Repairs What maintenance and repairs were necessary since your last trip? (i) On the hull (e.g. paint, caulking, etc.)? How much (total) did the materials cost? How much did the labor cost? How much of these costs was paid by the boats owner? How much was paid by the crew? By others? Who supplied the labor? Owner? - Crew? Family? How many hours or days were involved? (ii) Repairs on motor (filters, spark plugs, etc.)?

How much did the materials cost? How much did the labor cost? How much did the owner of the motor pay? How much was paid by the crew? How much was paid by others? Who supplied the labor? Owner? - Crew? Family? How many hours or days were involved? (iii) Repairs on the gear (like mending, replacing hooks, floats etc.)? How much did the materials cost? How much did the labor cost? How much was paid by the owner of the gear? How much was paid by the crew? Others? Who supplied the labor? Owner? Crew? Family? How many hour j were involved? (iv) Because of fhese repairs were any days of fishing lost since the preceding trip? How many da) s? (v) Did you lose any equipment this trip? What? What is the present value? What will it cost to replace it? How much of this replacement cost will be paid by the owner? How much will be paid by the crew? - Others? (vi) Have you bought any new equipment since the last trip? What? How much did it cost? How muc!! was paid by the owner? How much was paid by the crew? Others? (vii) Was any of your equipment damaged this trip? What?

114 What will it cost to repair? Materials Labor Who will pay for it? Who will supply Lthe labor?

APPENDIX D Village Based Interview: Fixed Cost Questions

I. Service Costs What annual or monthly expenses do you have for the following: Mooring Fee Monthly Annual Guard/Protection Monthly Annual Membership Monthly Annual Accounting Monthly Annual Legal (Insurance) Monthly Annual Other Monthly Annal Other Monthly Annual _ Other Monthly Annual

I. Crew Coats Do you provide housing, food, education fees etc., to your crew or their families? If so what? Monthly Annual Monthly Annual Monthly Annual Monthly Annual

I1. Invertiment Costs Hull Type: Length: Width: What is the capacity (maximum) of fish that the boat (hull) can carry? How old is the hull? How much did it cost you? When did you buy it? What would be its value if you sold it now? How much did it cost hull) when it was new? How much would it cost to buy a new hull equal to this one? How much longer will the hull last - working as it is? Have you made any major improvements or repairs in the hull in the recent past? (1-2 years) When? What? How much did it cost? Did you get a loan to buy the hull? (If he says yes, continue): (If he says no, go to motor). How much was the loan for? When did you receive it? month year How much did you actually receive? Who is the creditor? Bank? Which bank? Family? Friend? Money Lender? - Financer? Other? How much do you pay monthly against this debt? Interest Principal Is this monthly payment decreasing? Remainin3 Constant? or increasing? What rate of interest are your paying? How many payments do you still have to make? Did you have to put up any collateral as insurance on the loan? What did you use as collateral? What is its value?

115 Motor What brand is the motor? Horse Power? What does it use? Gasoline or diesel fuel? How old is the motor? When did you buy it? How much did you pay for it? How much would you get if you sold it today? What did it cost when it was new? How much would it cost to buy the same new motor today? How much longer will thiA motor last? When? Have you made any major improvements or repairs in the motor in the recent past? (1-2 years) What? How much did it cost? Did you get a loan for this motor? How much was the loan for? receive? When did you receive it? Month - Year How much did you actually Who is the creditor? Bank? Which Bank? Family? Friend? Money lender? - Financier? Other? Total What is the monthly payment on this loan? Interest Principal Increasing? Is the monthly payment getting smaller? Staying constant? What interest rate are you paying? How many more payments remain? Did you put up any collateral for the loan? What did you offer? For what value?

Gear What kinds of gear do you have? What size? Gear Type What kind? How many? A B B

How old is each gear A B C What is its value if you sold it today? A B C How much did it cost you? A B C When did you buy it? A B C What did it cost when it was new? A B C How much would it cost to replace it with new gear of the same type(s)? A - - B C How much longer will it last? A B C Did you get a loan to buy the gear? A B C How much was the loan for? A B C When? Month & Year? A B C

116 How much did you actually receive for each piece of gear? A B C Who is the creditor? A B C How much in total do you pay on these loans monthly? How much of this is interest? Principal? Are tlkese payments going down? Staying constant? Getting larger? What interest rate are you paying on these loans? How many more monthly payments must you make? Did you put up any collateral for these loans? What was the collateral and its value?

Other related to fishing activity? Do you have (own) other equipment: for iPxample, vehicles, buildings, docks, i.e. any equipment Do you have any other major debts? For what? What is the size of the debt? Owed to whom? How much do you pay monthly? Are these payments going down? Staying constant? What interest rate are you paying? How many more payments must you make? Did you put up any collateral for these loans? For what value?

117 BIBLIOGRAPHY

ALEXANDER, P. BRANDER, K. 1975. Innovation in a cultural vacuum: 1975. Guidelines for collection and com­ the mechanization of Sri Lanka fisher- pilation of fishery statistics. FAO Fish. ies. Human Organization 34:333-344. Tech. Paper 148. ANDERSON, L.G. (ed) BROMLEY, D.W. and R.C. BISHOP 1981. Economic Analysis for Fisheries 1977. From Economic Theory to Fisher­ Management Plans. Ann Arbor Science ies Policy: Conceptual Problems and Publishers: Ann Arbor. Management Proscriptions. In Ander­ son (ed.), Economic Impacts of ANDERSON, L.G. (ed) Extended Fisheries Jurisdiction. Ann 1977. Economic Impacts of extended Arbor Science Publications: Ann Arbor. Fisheries Jurisdiction. Ann Arbor Science Publishers: Ann Arbor, BROMLEY, D.W. Michigan. 1977. Distributional Implications of the Extended Economic Zone: Some Policy ANDERSON, L.G. and Research Issues in the Fishery. 1977. The Economics of Fisheries Man- American Juarnal of Agricultural Eco­ agement. John Hopkins University nomics. 59:887-892. Press: Baltimore and London. BROTHERS, E.B. 198C. Age aid growth studies on tropi­ BANERJI, S.K. cal fishes. In Roedel, P. and S. Saila (eds), Stock Assessment for Tropical sample1974. Frame survey surveys designs andfor theassociated assess- Small-Scale Fisheries. ICMRD: Univ. of Kingston. ment of marine fish landings. Indian R.I., Ocean Fish. Comm., U.N. Develop. Prog., IOFC/DEV/74/39. BURCZYNSKI, J. the use of sonar BAZIGOS, G.P. 1979. Introduction to of fisheries statistical systems for estimating fish biomass. 1974. The design 191. surveys - inland waters. FAO Fish. FAO Fish. Tech. Paper Tech. Paper 1.33. CAMPLEMAN, G.

BEVERTON, R.J.H. and S.J. HOLT. 1976. Manual on the Identification and 1956. A review of methods for estimat- Preparation of Fishery Investment Pro­ ing mortality rates in exploited fish pop- jects. FAO Fisheries Technical Paper ulations, with special reference to N4o. 149 (FIII/T149): Rome. sources of bias in catch sampling. Rapp. P.-v. Reun. Cons.67.83.1981. int. Explor. Mer 140(1): CASLEY, DataD.J. andCollection D.A. LURY in Developing 67-83. Countries. Oxford University Press: BEVERTON, R.J.H. and S.J. HOLT. Oxford. populations.1957. On the dynamicsU.K. Min. of exploitedAgric. Fish. fish CHAKRABORTY, D. and H. WHEELAND Food, Fishery Investigations (Ser. 2), 19. 1979. Report on Training Course in Fishery Statistics (SCS/GEN/79/23). BEVERTON, R.J.H. and S.J. HOLT. South China Sea Fisheries Develop­ 1959. A review ofthe lifespans and mor- ment and Coordinating Programme. tality rates of fish in nature, and their P.O. Box 1184, M.C.C. Makati Metro relation to growth and other physical Manila, Philippines. characteristics. In Wolstenholme, M. (eds.), The life- CHEIN, I. G.E.W. &O'Connor, In spans of animals. Ciba Found. Colloq. 1976. An Introduction to Sampling. 5:147-177. C. Selltiz, et al., (eds.), Research on Ageing Methods in Social Relations. Holt, Rine­ BEVERTON, R.J.H. and S.J. HOLT. hart and Winston: New York. 1966. Menual of methods for fish stock assessment. Part II - Tables of yield CLARK, COLIN W. FAO Fish. Tech. Paper 38. 1976. Mathematical Bioeconomics. functions. John Wiley &Sons: New York. 119 Previous Page Blank COCHRANE, C. ENGSTROM, J.E. 1979. The Cultural Appraisal of Devel- 1974. Preparation of Fishery Invest­ opment Projects. Prager Publishers: ment Projects. FAO. (IPFC/74/SYM/26). New York. Rome. COMITINI, S. EPLER, B. H.C. LAMPE, and J.G. 1975. Views of an Economist in the Aca- SUTINEN demic World, In Philip M. Rodel (ed), 1980. Fresh Fish Retail Marketing in Optimum Sustainable Yield as a Con- Costa Rica. In Sutinen, J. and R. Pol­ cept in Fisheries Management, Special lnac (eds), Small Scale Fisheries in Cen­ Publication No. 9. American Fisheries tral America: Acquiring Information for Society: Washington, DC. Decision Making, ICMRD: Univ. of R.I., Kingston. COPES, P. 1972. Factor Rcnts, Sole Ownership and EPPLE, G.M. the Optimal Level of Fisheries Exploita- 1977. Technological change in a Gren­ tion. The Manchester School ofEconom- ada, W.I., fishery, 1950-1970. In M.E. ics and Social Sciences. 41 (June): Smith (ed.), Those Who Live from the 145-163. Sea. West Publishing Co.: San COSLIT, R.Y., LAMPE, H.C. and Francisco. SUTINEN, J.G. 1980. Characteristics of Demand for FIRTH, R. Fresh Fish in Guatamala. In Sutinen, J. 1966. Malay Fishermen, 2d ed. Archon and R. Pollnac (eds), Small Scale Fisher- Books: Hamden, Conn. ies in Central America: Acquiring Infor­ mation for Decision Making. ICMRD: FISCHER, W. (ed.) Univ. of R.I., Kingston. 1978. FAQ species identification sheets CRUTCHFIELD, J.A. for tshery purposes: Western Central 1975. An Economic View of Optimum Atlantic (Fishing Area 31). FAO: Rome, Sustainable Yield. in Philip M. Roedel Vols. I-VI. (ed), Optimum Sustainable Yield as a Concept in Fisheries Management, Spe- FOOD AND AGRICULTURE ORGANI- cial Publication No. 9. American Fisher- ZATION OF THE UNITED NATIONS. ies Society: Washington, DC. sea.1972. FAO: Atlas Rome. of the living resources of the DIWAN, R.K. and D. LIVINGSTON FOOD AND AGRICULTURE ORGANI- 1979. Alternative Development Strate- ZATION OF THE UNITED NATIONS. gies and Appropriate Technology: 1976. Monitoring of fish stock abun­ Science Policy for an Equitable World dance: the use of catch and effort data. Order. Pergamon Press: New York, FAQ Fish. Tech. Paper 155. Oxford, Sidney, Paris, Toronto. DULOY, J.H. and R.D. NORTON FOSTER, G.M. 1973. CHAC, A Progkamming Model of 1973. Traditional Societies and Tvchno­ Mexico Agriculture. In L. Goreaux and logical Change, 2d ed. University of A. Manne (eds), Multilevel Planning: California Press: Berkeley. Case Studies in Mexico. North Holland Publishing Co.: Amsterdam. FOX, W.W., JR. 1970. An exponential surplus-yield model for optimizing exploited fish pop­ EMMERSON, D.K. Soc. 99(1): 1980. Rethinking Artisanal Fisheries ulations. Trans. Amer. Fish. Development: Western Concepts, Asian 80-E8. Experiences. Staff Working Paper No. World Bank: Washington, DC. FRASER, T.M. 423. The 1966. Fishermen of South Thailand. EMMERSON, D.K. Holt, Rinehart and Winston: New York. 1975. Orders of meaning: understand­ ing change in a fishing comunity in FREUND, J.E. Indonesia. Paper presented at the 1960. Modern Elementary Statistics annual meeting of the American Politi- (2nd edition). Prentice-Hall, Inc.: Engle­ cal Science Association, San Francisco. wood Cliff, N.J.

120 GARROD, D.J. HOLT, S.J. 1969. Empirical assessments of cat- 1962. The application of comparative ch/effort relationships in the North population studies to fisheries biology Atlantic cod stocks. Res. Bull. Intern. - an exploration. In Le Cren, E.D. & Comm. N.W. Atl. Fish. 6: 26-34. Holdgate, M.W. (eds.), The exploitation of natural animal populations. British GELLES, R. Ecol. Soc. Symp. No. 2: 51-71. 1974. The Violent Home: A Study of Physical Aggression between Hus- HUETH, D.L. and NORTON, V.J. bands and Wives. Sage Publications: 1977. Distributional Implications of the Beverly Hills. Extended Economic Zone: Some Policy and Research Issues: Discussion. Amer­ GI'1T'INGER, J.P. ican Journal of Agricultural Economics. 1972. Economic Analysis of Agricultu- 59: 893-897. ral Projects. The World Bank, John Hop­ kins University Press: Baltimore and HYSLOP, E.J. London. 1980. Stomach contents analysis - a review of methods and their application. GORDON, H.S. J. Fish. Biol. 17: 411-429. 1954. The Economic Theory of a Com­ mon Property Resource. J. Political JONES, R. Economy. 62 (April): 124-142. 1966. Manual of methods for fish stock assessment. Part IV - Marking. FAO G"TLLAND, J.A. Fish. Tech. Paper 51, Suppl. 1. 1966. Manual of sampling and statisti­ cal methiods for fisheries biology. Part I. JONES, R. Sampling methods. FAO Manuals in 1976. The use of marking data in fish Fish. Sci. No. 3. popul"ti.xi analysis. FAO Fish. Tech. Paper 153. GULLAND, J.A. 1969. Manual of methods for fish stock JONES, R. assessment. Part I. Fish population 1979. Materials and methods used ini analysis. FAO Manuals in Fish. Sci. No. marking experiments in fishers 4. research. FAO Fish. Tech. Paper 190. GULLAND, J.A. KUTTY, M.K. 1971. The fish resources of the ocean. 1970. The estimation of optimum condi­ Fishing News (Books) Ltd.: West tions for regulating a fishery when Byfleet, Surrey, England. growth and mortality rates are unknown. Proc. Indian Natl. Sci. Acad. HARPER, W.M., TWEETEN, L.G. Vol. 36, Part B,No. 1: 21-32. 1977. Sociopsychological Measure of LAXENAIRE, M. Quality of Rural Life: A Proxy for Meas- 1973. Nociones de Contabilidad y Ges uring the Marginal Utility of Income. tion para Marinos Pescadores: Manual American Journal of Agricultural Eco- del Divulgador. FAO(FIEE/T27). nomics. 59 (5). Rome.

A.W.HENDERSON, KUDHONGANIA.MAKTDJ H.F. R.A. RYDER, and MACKEFT, D.J. 1973. Assessing fishery potentials of 1973. Manual of methods for lakes and fisheries reservoirs. J. Fish. Reas. Bd. resource survey and appraisal. Part III Canada 30: 2000-2009. - Siandard methods and techniques for demersal fisheries resource surveys. H.E.W. (Dept. Health, Education, & FAO Fish. Tech. Paper 124. Welfare) MUNRO,J.L 1973. Work in America. M.I.T. Press: 1980. Stock Assessment Models: Applic­ Cambridge. ability and Utility in Tropical Small- Scale Fisheries. In Roedel, P. and S. HOGBEN, L. Saila (eds), Stock Assessment for Tropi­ 1968. Statistical Theory. Norton Press: cal Small-Scale Fisheries. ICMRD: New York. Univ. of R.I., Kingston.

121 MUNRO, J.L. and THOMPSON, R. PAULY, D. 1973. Th- Jamaican Fishing Industry, 1980b. On the interrelationships the Area Investigated and the Objec- between natural mortality, gr-wth tives and Methodology of the parameters and mean environmental ODA/UWI Fisheries Ecology Research temperature in 175 fish stocks. J. Cons. Project. Research Report 3. Zoology int. Explor. Mer. 39(2): 175-192. Department, University of West Indies, Kingston, Jamaica. PEARSE, P.H. 1979. Symposium on Policies for Eco­ NARKSWASDI, U. nomic Rationalization of Commercial 1967. An Evaluation of the Kuala Fisheries. J. Fisheries Research Board Linggi Fishermen's Cooperative Credit of Canada. 36(7). and Marketing Society, Ltd. Ministry of Agriculture and Cooperatives: Kuala PELLA, J.J. and P.K. TOMLINSON Lumpur. 1969. A generalized stock production Inter-Amer. Trop. Tuna N.H., HULL, CH., JENKINS, J.G., model. Bull. NIE, Comm. 13(3): 419-496. STEINBRENNER, K., BENT, D.H. 1974. Statistical Package for the Social Sciences, 2d ed. McGraw-Hill Book Com- POLLNAC, R.B. pany: New York. 1982 Sociocultural aspects oftechnolog­ ical and institutional change among NORR, K. small-scale fishermen. In J.R. Maiolo 1972. A South Indian fishing village in and M.K. Orbach, (eds), Modernization comparative perspective. Ph.D. disser- and Marine Fisheries Policy. Ann Arbor tation, Univ. of Michigan. Science Publishers: Ann Arbor, Mich.

OVENDED, A.E. POLLNAC, R.B. 1961. Cost and Earnings Investigations 1981. Sociocultural Aspects of Develop­ of Primary Fishing Enterprises: a Study ing Small-Scale Fisheries: Delivering of Concepts and Definitions. FAO. Services to the Poor. World Bank Staff Fisheries Study No. 10. Rome. Working Paper. No. 490.

PAULY, D. POLLNAC, R.B. 1978. A preliminary compilation of fish 1979. Cambio technologico y organiza­ length growth parameters. Berichte cion social entre los pescadores artesa­ Inst. f. Meereskunde (Kiel) 55. nales. In Actas, Seminario/Taller Sobre Desarrollo y Investigacion de los Recur­ PAULY, D. Region, Chile, size and temperature as sos Marinos de la Octava 1979. Gill Univ. de Concep­ factors in fish growth: A gen- V.A. Gallardo (ed.), governing Concepcion. eralization of von Bertalanffy's growth cion.: formula. Berichte Inst. f. Meereskunde (Kiel) 63. POLLNAC, R.B. 1978. Sociocultural aspects of Techno­ PAULY, D. logical and Institutional Change 1979. Theory and Management of Tropi- among Small-Scale Fishermen. Anthro­ cal Multispecies Stocks: A review, with pology Working Paper No. 22, Univ. of emphasis on the southeast Asian demer- Rhode Island. sal fisheries. ICLARM Studies and Reviews. No. 1. International Center for POLLNAC, R.B. Living Aquatic Resources Manage- 1976. Continuity and Change in Marine ment: Manila. Fishing Communities. A State-of-Art Paper prepared for the U.S. Agency for PAULY, D. International Development. !980a. A new methodology for rapidly basic information on tropical acquiring J. POGGIE fish stocks: growth, mortality, and stock POLLNAC, R. and relationships. In Roedel, P. 1979. The Structure of Job Satisfaction recruitment Fishermen. and S. Saila (eds), Stock Assessment for Among New England Small Scale Fisheries. ICMRD: Anthropology Working Paper No. 31, Tropical Island. Univ. of R.I., Kingston. Univ. of Rhode

122 POLLNAC, R.B. and R. RUIZ-STOUT SABELLA, J.C. 1977. Small-scale fishermen's attitudes 1974. The Fishermen of Caleta San towards the occupation of fishing in the Pablo. Ph.D. dissertation, Cornell Univ. Republic of Panama. In R.B. Pollnac, Editor, Panamanian Small-Scale SABRI, J. Fishermen: Society, Culture, and 1977. Small-scale fisheries development Change. Marine Technical Report No. in peninsular Malaysia -problems and 44, University of Rhode Island, King. prospects. In (eds) B. Lockwood and K. ston, R.I. Ruddle, Small-Scale Fisheries Develop­ ment: Social Science Contributi,n, POLLNAC, R.B., and J. SUTINEN East-West Center Press: Honolulu. 1980. Economic, social, and cultural aspects of stock asessment for tropical SAMUELSON, P.A. small-scale fisheries. In Roedel, P. and 1973. Economics. 9th ed. McGraw-Hill S. Saila, eds. Stock Assessment for Trop- Book Company: New York. ical Small-Scale Fisheries. ICMRD: SASSONE, P.G. and SCHAEFFER, W.A. Univ. of R.I., Kingston. 1978. Cost-Benefit Analysis: A Hand­ POPE, J. book. Academic Press: New York. 1979. Stock Assessment in Multispecies SAVILLE, A. Fisheries, with special reference to the 1977. Survey methods of appraising trawl fishery in the Gulf of Thailand. 1977. surces O F ising Fisheries Technical Paper SCS/DEV/ fishery resources. FAO Fish. Tech. 79/19. South China Sea Fisheries Devel- Paper 171. opment Coordinating Programme: SCHAEFER, M.B. Manila. 1954. Some aspects of the dynamics of the fishery for yellowfin tuna in the east­ RETTIG, R.B., GINTER, J.C. (eds) ern tropical Pacific Ocean. Bull. Inter- 178. Limited Entry as a Fishery Man- Amer. Trop. Tuna Comm. 1: 26-56. Nationalagement Tool,Conference Proceedings to Consider of a SCHANK, M., SUTINEN, .J.G. Limited Entry as a Tool in Fishery Man- 1980. An Economic Analysis of Fish agement. University of Washington Marketing in El Salvador. In Sutinen J. Presetl ndersindon.Washiand R. Pollnac (eds), Small Scale Fisher­ Press: Seattle and London. ies in Central America: Acquiring Infor­ RICKER, W.E. mation for Decision Making. ICMRD: 1975. Computation and interpretation Univ. of R.I., Kingston. of biological statistics of fish popula- SCHEID, A., SUTINEN, J.G. tions. Bull. Fish. Res. Bd. Canada 191. 1980. The Structure and Performance oi RODEL, P.M. and SAILA, S.B. Wholesale Marketing of Finfish in 1980. Stock Assessment for Small-Srcle Costa Rica. In Sutinen, J. and R. Pol­ Fisheries. ICMRD, University of Rhode lnac (eds), Small Scale Fisheries in Cen­ Island: Kingston, R.I. tral America: Acquiring Information for Decision Making. ICMRD: Univ. ofR.I., ROGERS, E.M., and F. SHOEMAKER Kingston. 1971. Communication of Innovations, 3d ed. New Yo'k: The Free Press. SCOTT, A. 1977. Commentary on Chapter 15. In j.iDER, R.A. Lee G. Anderson (ed), Economic 1978. Fish yield assessment .f !arge Impacts of Extended Fisheries Jurisdic­ lakes and reservoirs - prelude to man- tion, Ann Arbor Science Publishers: agement. In Gerking, S.D. (ed.) Ecology Ann Arbor, Michigan. offreshwater fish production, Blackwell Sci., Oxford: 403-423. SEIBEL, H.D., and A. MASSING 1974. Traditional Organizations and RYDER, R.A., S.R. KERR, K.H. LOFTUS Economic Development. Praeger Pub­ and H.A. REGIER lishers: New York. 1974. The morphoedaphic index, a fish yield estimator - review and evalua- SELVIN, AH.C. tion. J. Fish. Res. Bd. Canada 31: 663- 1957. A Critique of Tests of Significance 688. in Survey Research. American Sociolo­ gical Review 22: 519-27.

123 SELVIN, F.J. YAP, C.L. 1975. The Fisherman's Business Guide. 1977. Trawling: its impact on employ­ International Marine Publishing Co.: ment and resource use on the west coast Camden, Maine. ofpeninsular Malaysia. In (eds) B. Lock­ wood and K. Ruddle, Small-Scale SMITH, IAN R. Fisheries Development: Social Science 1979. A Research Framework for Tradi- Contribution, East-West Center Press: tional Fisheries. ICLARM Studies and Honolulu. Ro-views No. 2. International Center for Living Aquatic Resource Management: ZIONTS, S. Manila. 1973. Linear and Integer Programming. SMITH, P.E. and S.L. RICHARDSON Prentice-Hall, Inc.: Englewood Cliffs, 1977. Standard techniques for pelagic New Jersey. fish egg and larva surveys. FAO Fish. Tech. Paper 175. STEVENSON, D.K. 1980. Use of length-frequency data to estimate growth and mortality rates for species exploited by tropical small-scale fisheries in Puerto Rico and Costa Rica. In Roedel, P. and S. Saila (eds), Stock Assessment for Tropical Small-Scale Fisheries. ICMRD: Univ. of R.I., Kingston. SUTINEN, J., and POLLNAC, R.B. (eds.) 1980. Small-Scale Fisheries in Central America: Acquiring information for Decision Making. International Center ior Marine Resource Development, Uni­ versity of Rhode Island: Kingston, R.I.

THOMAS, D.H. 1976. Figuring Anthropology. Holt, Rinehart and Winston: New York.

UNITED STATES DEPARTMENT OF COMMERCE Tide tables. National Oceanic and Atmos. Admin., National Ocean Sur­ vey, Rockville, Md. UNITED STATES NATIONAL ACADEMY OF SCIENCES 1969. Recommended procedures for measuAng the productivity of plankton standing stock and related oceanic prop­ erties. U.S. Gov't Pr!.nting Office, Wash., DC.

VOLLENWEIDER, R.A. (ed.) 1969. A manual on methods for measur­ ing primary production in aquatic envir­ onments. IBP Handbook No. 12, F.A. Davis: Philadelphia.

WELCOMME, R.L. 1979. Fishery management in large riv­ ers. FAO Fish. Tech. Paper 194.

124