DYNAMIQUE ET USAGE DES RESSOURCES EN SARDINELLES DE L’UPWELLINGCôTIER DU GHANA ET DELA CôTE-D’IVOIRE DYNMICS AND USE OF SARDINELLA RESOURCES FROM UPWELLING OFF GHANA AND IVORY COAST
Actes du colloque DUSRU, Accra, du 5 au 8 octobre 1993 Acts of DUSRU meeting, Accra, october 5 8 t h 1993
.-
DYNAMIQUE ET USAGE DES RESSOURCESEN SARDINELLES DE L'UPWELLING CôTIER DU GHANA ET DE LA CÔTE-D'IVOIRE DYNAMICS AND USE OF SADINELLA RESOURCES FROM UPWELLING OFF GHANA AND IVORY COAST
Éditeurs scientifiquesl Scientifics editors
François Xavier BARD,Kwame A. KORANTEG ._. "
-
,,
...,,.. ... , ;< 3!. ~.___ . *
.
1,
I
6
\+A!
"
.,&.:
'7%. 'i ,
":
._.,
":,,!
&$[
))::-:;$,>, .'.,;,:'
'1
8
1
L'INSTITUT FRANCAIS OERECHERCHESCIENTIFIOUE POUR LE DiVELOPPEMENT EN COOP&iATION
CENTRE
DE RECHERCHES OC&NOtOGlOUES DE COTE-O'IVOIRE
FISHERY RESEARCH UNIT BRANCH OF GHANA
ORSTOM Éditions INSTITUT FRANçAIS DE RECHERCHE SCIENTIFIQUE POUR LE DÉVELOPPEMENT EN COOPÉRATION
Collection COLLOQUES et SÉMINAIRES PARIS 1995
Le grand programme(( Sardinelies )) a kt6 men6 avec le soutien du ministgre franpis de la Cooperation.
La loi du 11 mars 1957 n'autorisant, aux termes des alinbas 2 et 3 de l'article 41, d'une part, que les ((copies ou reproductionsstrictement rbservees B l'usage prive du copiste et non destinees B une utilisation collective)) et, d'autre part, que les analyses et les courtes citations dans un but d'exemple et d'illustration, ((toutereprbsentation oureproduction integrale, ou partielle, faite sans le consentement de l'auteur ou de ses ayants droit ou ayants cause, est illicite)) (alinbaler de l'article 40). Cette reprbsentation ou reproduction, par quelque proc6db que ce soit, constituerait donc une contrefaçon sanctionnbe par les articles 425 et suivants du Codephal.
ISSN : 0767-2896 ISBN 2-7099-1258-2
O ORSTOM Éditions 1995
A M d a m aérèse TU~QCL?, sans qui, rien du fiogramm Sardine& n aurait éte" possi6ce. A la mémoire de Jean Louis Durand qui fut un 6on
compagnon.
FOREWORD This book is the product of three years of joint research effort by scientists of the Institut Francais de Recherche Scientifiques pour le Developpement en Cooperation (ORSTOM) of France, Fishery Research and Utilization Branch (FRUB) of Ghana, and Centre de Recherches Oceanologiques (CRCI) of Cote d’Ivoire. The project aimed at understanding the proliferation of sardinellas in the ‘coastal ecosystem off Cote d’Ivoire and Ghana and which was dubbed the ((SardinellaPuzzle>>.This puzzle corresponds to the surprising observation made at the end of 1987, that the apparent abundance of Sardinella aurita had steadily increased since 1984. After some perplexities about this phenomenon, some fisheries scientists of Ghana and Cote d’Ivoire, with the assistance of ORSTOM, decided to study the possible origin and consequences of the apparent proliferation of the sardinellas. A programme, entitled >of such species (e.g. the Japmese sardine by 1985) could also occur without simple explanation. The coastal pelagic resources of the Côte d’Ivoire-Ghana region ,are of special interest because the stocks involved are small, have been studiedfor a long time and are of great socio-economic importancefor the coastal populations. A lot of scientists have been involvedin biological studies, fisheries and environmental components of the coastal pelagic ecosystemof C6te d’Ivoire and Ghanasince the 1960”s. Moreover a dramatic increaseof the abundance of Sardinella aurita has been observed in the 1980 S . ) . This phenomenon, dubbed is being studied intensively by Ghanaian, Ivorian and French scientists, since 1988. i By mid 1990 it was decided to increase the research effort, particularly through an intensivejoint program submittedas &rand Programme Sardinelle>> (GPS) to various potential donors.It was accepted by ORSTOM and submitted to the French Ministryof Co-operation for funding. From 1991 to 1993 this France-Ghana-Côte d’Ivoire joint Sardinella research programme (GPS), has been in operation. Various results described herewith have been achieved. This led to the end of the first phase of the GPS. An intermediate meeting, dedicatedto an evaluationof the results of this first phase, and to the preparation of a program for a possible second phase, was decided. Meanwhile, Mr. Levallois, ORSTOM President, having been introducedto the GPS, during his visit to Ghana, in May 1992, suggested thatthe intermediate GPS meeting shouldbe held in Accra, takinginto account the potential interest for Ghanaian economic and political authorities and privateof users the resource.
30
Therefore this DUSRU meeting is to be jointly organised by the Fisheries Department of Ghana ( Ministry of Food & Agriculture), and ORSTOM.FA0 and French Embassy in Accra indicated their willingness to collaborate.
2 . AIMS AND OBJECTIVESOF THE MEETING This meeting would be an opportunity to present to the world scientific community, and also the economic and political authorities of Ghana and Côte d’Ivoire, the situation of the knowledge of this ecosystem, the resources and their utilisation by man. Different but complementary parts are evident, namely: the fishes and their biology, the fisheries and socio-economics and the environment and its variability which affects boththe fishes and thefisheries. This meetingis intended to be an occasion to the open discussion toal1 parties concerned and have the view of some of the well-known experts on such problems. Therefore, the meeting is aimed at presenting knowledge acquired in over 25 years of joint research on the pelagic resourcesof the Gulf of Guinea, especially on the Sardinella, and place thisparticular situation in the appropriate perspective in the world. For this purpose, the meeting willbe connected withthe Climate and Eastern Ocean Systems (CEOS)programwhich is a collaborative study addressing potential effects of global versus local climate change on the populations of the highly productive ecosystems of eastern ocean upwelling regions, and on directly associated economic, cultural and ecological issues. 3. FORM OF THE MEETING
This meeting will be heldin English andit could be divided into twoparts : 1. A papers presentation sessionby invited speakers which will present the general problem according tothe following frameworkof basic questions : - Changes in coastal pelagic populations during the last decade, off GhanaCôte d’Ivoire ;artefacts or real changes?. - Fisheries impact and/or environmental changes?. - 1s Ghana-Côte d’Ivoire upwelling an exception among other upwelling ecosystems ?. - What can we l e m from other upwelling ecosystems and their related fisheries ?. - What future researches are needed ?. These items concern the questions related the to Côte d’Ivoire - Ghana region but also those involved in the CEOS program. 2. A session of workshops which would discuss particular points. Two workshops are proposed. They will be organised asa comparative approach in order to study the basic case of Ghana-Côte d’Ivoire, and compare it with other cases in the rest of the ecosystems of eastern ocean upwelling regions.The two themes will be :
31
- Environrnental variability ; impact on coastal rnarinepelagic populations and resources, particularlySardinella aurita. - Economics and management policies of coastal pelagicfisheries in Africa and in the world.
A book will be prepared and edited jointlyby ORSTOM and FRUB. It will include invited lectures in English or French , reports of the workshops, recommendations and conclusions. Scientific Editors will be members of the steering comrnittee. A data base includingthe information retrieved, gathered and cheeked during the first phase of GPS willbe made available to participants, on diskette. A research programof the second phaseof the GrandProgramme Sardinelle (1 994- 1996) will be preparedby end of 1993 in order to seelr for potential donors and joint political endowment of Ghana andCôte d’Ivoire governments. It will be based on the recornmendations madeby the participants for future research on environment, fisheries, trade, sociology and economics linked topelagic small resources of the Ivorio-Ghanaian upwelling ecosystem.
Fimt day :October, 318, 1993.
Morning : . Intrsduetisn of Chairman
e Sardinellla Prsblem Dr. Martin A. Mensah, Deputy Director of Fisheries e need ffor esoperative research on the sardinella resources in the estern Gulf of Guinea. Representative in Cote d’Ivoire TOM, Mr. Levallois Address byM. B’Ambassadeur de France, M. Brochenin Opening address The Deputy Minister for Food and Agriculture, Major (Retired) E. Tetteh Second Day :October 6 th, 1993.
Al1 the day long at FA0 meeting room : Presentation of Papers classified according to the three following categories : - Physical processes and environmental conditions. - Environment impact on biology, ecology and fisheries - Socio-economics and management policies. 32
Third day October, 7th, 1993.
Al1 ‘the day long, Workshops in FA0 building. Fourth Day October, Sth, 1993
Morning a t F A 0 meeting room: Presentation of workshop reports, conclusions and recommendations. Afternoon : Closure : Closing addresses by Chairmanof meeting ,and Director of Fisheries.
ANNEX 2
LIST OF SPEAKERS AND PAPERS Angora A. Fofana S. (Université d’Abidjan, Côte d’Ivoire). CoastalSea Surface Temperature as detected by the METEOSAT satellite and received at the University of Abidjan. Bakun A. (FAO, Rome, Italy). Global changes and the impact on the Gulf of Guinea and other upwelling areas. Bard F.X. (ORSTOM, Abidjan, Côte d’Ivoire). Predation by tropical tunas on Sardinella. Possible induced changes in tuna populationsby increase of Sardinella abundance. Binet D. (ORSTOM,Nantes, France). Hypothesis accountingfor the variability of Sardinella abundance inthe Northern Gulf of Guinea. The Chikhi L., Bonhomme F and Agnese J.F. (ORSTOM, Montpellier, France). search for genetic differentiation of two sardine species (S aurita, S maderensis). Cury P.(ORSTOMREG, Monterey, USA). Comparison of the Ivoiro-Ghanaian fisheries dynamics with other upwelling ecosystems. Durand M. H. (ORSTOM, Bondy, France). Mode de formation des prix des espèces pélagiques et enjeux d’exploitation pour les zones d’upwelling faiblement productives. Ferraris J., Koranteng K.A. (ORSTOM, Dakar, SenegalRRUB, Tema, Ghana). Statistical analysis of canoe fisheries data in Ghana with particular reference to Sardinella. Gbaguidi A. Pêcherieartisanale des Sardinelles auBénin. Konan J. (CRO, Abidjan,Côte d’Ivoire). Ivorian canoe fisheryfor Sardinella. Koranteng K.A. (FRUB, Tema, Ghana). The Ghanaian canoe fishery for Sardinella. 33
Kouakou Koffi J. (Direction des PCches, Abidjan, Côte d’Ivoire). Legalaspects in the management of the resources of small pelagic of the IvoiroGhanaian ecosystem. Kwei E.A. ,Koranteng, K.A. and Bannerman,P.O. (Tema, Ghana). TunaBait boat baiting time and the availabilityof anchovy in Ghanaian waters. and larvae Mensah M.A. (%;RUB,Tema, Ghana). Availability of Sardinella eggs off Tema during the period1969 - 4 992.
Mensah MA. (FRUB,Tema, Ghana). Occurrence of zoo plankton off Tema during the period 1969 - 1992.
N’dah M. (CRO) Age detemination and the hypothesis of annual double cycle of growth in Sardinella aurita (Val). Panfili J. (ORSTOM Brest). Use of otoliths for stock discrimination and qge estimation in Sardinella : proposa1 for a regional research project in WesternAfrica. Ecological importanceof’the Ivorian Pezennec O. (ORSTOM, Abidjan, Côte d’Ivoire). and Ghanaian minor upwellingseason. Roy C. ( O R S T O W ~ GMonterey, , USA). Ghana-Ivorian upwelling. Dynamics and environmental changes. Samba A. (CRODT, Dakar, Senegal).Senegalese canoe fishery for Sardinella. Seini, A. WAYO (ISSER, University of Ghana, Legon). Economics of canoe fisheries in Ghana.
Tettey E. (INFOPECHE, Abidjan, Côte d’Ivoire). West Africa Markets for sma1P pelagics.
’~
Tettey E., Koranteng K.A. (INFOPECHE, Abidjan, Côte d’Ivoire/FRUB, Tema, Ghana). Sardinella market trends inthe West African sub-region.
’
34
ANNEX 3 TEXTS OF THE INTRODUCTORY ADDRESSES WELCOME ADDRESS DELIVERED BY DR A. MENSAH ON THE OCCASION OF THE OPENING OF THE SClENTlFlC MEETING ON THE DYNAMICS AND USES OF SARDINELLA RESOURCES FROM THE UPWELLING OFF GHANA AND COTE D’IVOIRE (DUSRU) Mr. Chairman, Honourable Minister of Food and Agriculture
1deem it an honour to be given the opportunity to welcome Our distinguished guests and invited scientists from far and near to this all-important scientific meeting on the sardinella resources off Ghana and Cote d’Ivoire.1believe that quite a good number of you Who have come from outside Ghana have already been to Ghana before and most likely, several times. For this reason the proverbi Ghanaian hospitality is nothing new to most of you, distinguished guests. To the very first time visitors to this country,you will surely find the broad smiles on the faces of most Ghanaians Who are al1 too willing to help you out of any difficulty whether in the streets, in the play houses, in the tourist places or the markets. Whenever you are convinced that there is a problem, the Ghanaian will smilingly assure you otherwise by telling you: fish enter the fishery, is not yet satisfactorily solved in the case S.ofaurita the inthe Western Gulf of Guinea. Stock-recruitment relationship in the context of the coastal upwelling has been examined byseveral fisheries scientists andoceanographers in the West African sub-region. However no satisfactory or conclusive results have beenobtained. 37
Meanwhile, it is only assumed that: a) a good upwelling enhances spawning and increases the chances of good recruitment, b) analysis of the lengths of fish (Le. the length of the fish from the tip of the head to the fork of the tail) indicate the age structure of the sardinella population (i.e. abundanceo f fish spawned in a particular year). c) the landings of the beach seine nets are crude indications of the strength of recruitment of young fish, including the sardinellas, into thefishery. This is so because the beach seine thrives largely on juvenile fish; a large very youngfish including proportion of landings of the beach seines is madeofup round and flat sardinellas. Thus, the problem of recruitment of the sardinellas into the fishery is still not understood. The Migration Pattern and the Question of Stocks of the Sardinellas :
The migration pattern of S. aurita had been studied in detail by several fisheries scientists in the western Gulf of Guinea and had come up witha model which was accepted. This modelconsiders that one stock of S. aurita exists off Ghana - Cote d’Ivoire. A large proportion of the stock rests off Central Ghana and at the onset 0 % the fishing season (July - September), the fish move or migrate shorewmds to Cote d’Ivoire and Western Ghana. They turn eastward as they hit the coast and continue in this direction to Togo and sometimesas far as Benin. The fish that successfully makethis trip, a distance of about 500 km (300 miles), return to their resting groundsabout the end of the season. Even though this model is still held valid, yet recent pieces of evidence indicate some departure fonn this patternof migration. Until the last seven fishing seasons or so, the pattern of sardinella fishing conformed with the migratory pattern as explained above.The fishemen followed the fish from the West to the east along the coast. Many years ago, when out-board motors were not very popular in Ghana, the canoe fishermen virtuallymoved camp tofollow the fish. Such migration of fishermen is very much reduced these days. Since 1987, it appears that not much ofthe fishcontinue this eastward journey pastthe Central Region of Ghana. Landingsof S. aurita in the Greater - Accra and Volta Regions are becoming quite less than before; also the inshore which vesselsmostly operate from Tema in the Greater-Accra Region have been recording their lowest catche Furthemore, in recent years,the catch in Cote d’Ivoire appears more spread out on the Ivorian coast al1 year round. How can these observations be explained, Mr. Chaiman ? Concerning the stock, thereare increasing doubts as to whetherone or more separate stock of the S . aurita exist off Cote d’Ivoire - Ghana. What do we think about this, Fellow scientists ? There is a second school of thought which maintains that there are &Q separate and independent stocksof the S. aurita in the Western Gulf ofGuinea one stock on the easternside of Cape Three Points (calledthe Ghana stock) and 38
the other on the western side of the cape (the Ivorianstock). Mr. Chairman,for how long shall we continue to speculate on such important issue ? Do the fish recognise the sovereignty of Our states ? The migration pattern of the flat sardinella is less understood. It appears that the flat sardinella is a coastal and relatively sedentary species and so its seasonal migration along the Coast is of a limited extent. It is obvious therefore that Our knowledge of the migratory pattern of the S. aurita and the number of stocks needs to be updated. Effective fishing Effort and the Problem of Modelling :
Several gears (fishing nets) are employed in the sardinella fishery : - ali poli, watsa, beach seines, purse seine nets and sometimes trawl nets are al1 used. In view of the different mesh sizes and manner of operation, these gears catch different sizes of the sardinellas; they have different efficiencies and effective fishing efforts. For this reason, it has proved very difficult to obtain aneffective fishing effort which could be applied in modelling. This is a challenge to us, fellow scientists. Mr. Chairman, distinguished guests, fellowscientists, the S. aurita fishery in the Western Gulf of Guinea, is a very important and target fishery to both the artisanal and semi-industrial fleets. For this reason the unpredictable high fluctuations in abundance and catch rates forthe past two decades or so affect the livelihood of the fishermen. It is very obvious from these unpredictably high fluctuations in abundance that the fishery is not as well understood aswas once thought. The estimationof the biomass of the round sardinella has remained a difficult problem. The high catches made since 1983cannot be reconciled with the low biomass estimated. The increasing catches especially that which was made in 1992 could also leada to disasterjust as increased catches in 1972 nearly collapsed the fishery. Furthermore, predictions about the future of such an unstable resource as theS. aurita in the-western Gulf of Guinea onthe basis ofOur present knowledge could bedifficult and inaccurate. With so much that is unknown with reasonable precision, there is the real need to raise the level of research on the sardinella resources and their environment in the Gulf of Guinea. Would it be too much to ask the coastal states in the sub-region to conduct regular acousticsurveys on the resources in order toobtain a better understanding of the state of the resources, recruitment, migrations, stocks, etc. ? The occasional surveysconducted in the sub-region by foreign vessels have just been stop-gaps. Mr. Chairman, distinguished guests, fellow scientists,, 1wish to believethat while enjoying such a congenial, peaceful and hospitable environment in Ghana, you would succeed, as a result of your expected fruitful discussions within the next four days, in at least, recommending the direction of future research that would lead to the solution of such important problems as alreadyexplained, in the environment, the biology and fishery of the sardinellas. 39
:1:
:>
*
Excellency, the Minister of Food And Agriculture, Monsieur le President de L’ORSTOM, Monsieur l’Ambassadeur de France FAB, Deputy Representativefor Africa Director of Fisheries, Mister Chairman, Ladies, Gentlemen
It is a great honourfor me to introduce our Co-operative program of scientific research which yielded this meeting. However, I beg your pardon for my ugly english, and 1 hope that the story of our efforts deserves your attention and is ’ anyway understandable for this most distinguished assembly. Small pelagic fish constitute the main fishery resources of the world. They represent between a third and a half of total world fishcatches, which presentely is around 50 millions tons. These small pelagics (that could be more exactly named small neritic fish) are used either for human consumption, generally at affordable price, or for fishmeal, being vitalfor any productionrequesting cheap animal proteins. Consequently the importanceof small pelagicheritie fisheries for the whole world economics is considerably high. However, the srnaIl pelagic fish resource chxacterised are by huge instability, probably due to their sensitivity to the environmental conditions. Larger resources of such species have experienced huge dramatic variations, either collapses w caused economic disasters for the regions concemed or strong fast recoveries. and The most famous example is certainly the story of the Peruvian anchoveta fishery, whose landings soared up to near 10 millions tons by the sixties, collapsed to nearly ni1 in the 1976 s, and presently are climbing again several to (six) millions tons per year. Anotherexample is the Japanese sardine which became suddenly very abundant by 1985, landings reaching4 millions tons per year. But fishery collapsed again in 1992. The coastal pelagic resources of the Côte d’Ivoire-Ghana region are of special interest because the stocks involved are small, have been studied for a long time and are of great socio-economic importancefor the coastal populations. A lot of scientists have been involved in biological studies,fisheries and environmental components of the coastal pelagic ecosystem of Côte d’Ivoire and Ghana since the 1960’s. One of the prominent result of these studies is a synthesis in the format of a blue book, achievedjointly by FRU, ORSTBM and@RBin 1976. A particular conclusionat this time, was that the Sardinella aurita (round sardinella) 40
resource had a production potential of no more than 100 O00 tons/year. This maximum potential having been reached in 1972, was followed by a strong depleted phase. This low level of the Sardinella aurita resource convinced the involved scientist that Ghana and Côte d’Ivoire were sharing a common resource, that dwindled for several years. Nevertheless a dramatic increase of the apparent abundance of Sardinella aurita has been observed in the eighties. This phenornenon, has been studied intensively by Ghanaian, Ivorian and French scientists, since December 1987. At this time, a meeting convened in C R 0 Abidjan discovered, with some delay,the importance of this phenomenon The huge level of the 1985-1987 catches of Sardinella aurita caused some surprise, and trying to understand what happened, this assembly of scientists was close to announcing for predicting the very near collapse of the stock. It would in away have been >, and fortunately we this refrained natural tendancy of scientists. Instead, we developeda strong Co-operation between FRU, CR0 with support of ORSTOM scientists working inCôte d’Ivoire. This Co-operation led to annual meetings either in Abidjan, or inTema, devoted to the scientific description of the level of the Ivorian-Ghanaian small pelagicheritic resource, particularlythe Sardinella aurita. It was indeed an improvement on past time where common meeting were held on a much less frequent rythm. As the increase in the apparent level of abundance of the resource stayed at a high level,by July 1990by the involved scientists decided to increase the research effort, particularly through an intensive joint program called the >that can be predicted with quite a high degree of land masses certainty is an increasedcontrast in temperature between the heated and theoceans during the spring-summer seasons (Bakun, 1990,1992). This is because the large heat storage capacity of water causes the oceanic temperature cycles to lag the terrestrial cycles. For example, in coastal upwelling regions the development of a > in relative abundance of sardines andanchovies were initiated in al1 four regions (California Current, Peru-Chile, N W Pacific, and Benguela Current) in the second half of the 1990s. Many other examples can be found of major marine population effects occurring during this same period of the mid-1970s to mid-1980s.For example, 65
this wasa period of increasing survival of Greenland halibut (Smebryakov,1992), a period of dramatic growth inthe lobster landings in eastern Canada (Pezzack, 1992), a period of large increase in Newfoundland spawning northern Cod stock (Paz and Larraiieta, 19921, etc. On the other hand, North Pacific Albacore tuna appear to have suffereda steep population decline during the period (FAO, 1992).
4. Potential Connections &O Thus long-term unidirectional trends, such as one rnight expect from the climate change mechanism discussed earlier in this paper, have not been the rule. Rather, shorter decadal-scale population have been more typieal. Moreover these swings have the appearance of being synchronized globally by the singular conditions of the decadal period fromthe mid-1960s to mid-1970s. Indeed,if one examines the varions series presented by B&un (1 990, 1992) or Roy (1990) it is clear that although multi-decadal coastal wind and upwelling increasesare clearly indicated, these trends have not been steady and unidirectional on the shorter(- decadal) scales. The most rapidincreases tended to occur fromthe mid-1950s to the early to mid-1970s. Then, thetrends in widely , separated regions,in both the northern and southern hemispheres and in boththe Atlantic and Pacific, appear to d i p overn to negative trends,or at least to level off. Then, near the mid-1980s there is often a suggestion of another infleetion point where the earlier long-tem upward trend seems to begin to be reestablished. The worldwide coherence of these trends and inflexion points is striking and quite extraordinary.This period (early to mid-1970s to mid-1980s) of temporary interruption of the longer term trends, is exactly the period of the remarkable global marine population episodes, including those involvingthe coastal pelagic fishes in the Gulf of Guinea. What could be the origin of these global decadal variations? The evidence suggests that it probably lies in the eoupled ocematmosphere system of the equatorial Pacific Ocean. The waters of the Pacific Bcean constitute the greatest mass ofheat storage capacity on earth. The Pacific is so large that its atmosphereas a whole is snly minimally subject to continental effects, and so is not nearly-so strongly forced into seasonal climatic regularity as are the atmospheres of smaller, more continentally-influenced oceans. The result is that the coupled ocem-atmosphere system of the Pacific that act to transmit ENSO-related effects to marine ecosystems distributed throughout the world’s oceans have become well documented and reasonably well understood bythe ocean scientific community. But underlying the ElNiiio eventsoccurring on the inter-annual scale,there appears to be a decadal-scale variation that appears as lower frequency modulations of both amplitude and frequency of interannual episodes.The period from the early to mid- 1970s to the mid- 1980s clearly onewas in whichthe Pacific Ocean was characterized by a rather continuous state of elevated El Niiio characteristics (Norton, et al., 1985; Wooster and Hollowed, in press). For example, if one looks at available time seriesof ocean properties and processes off Peru, one finds the interannual El Ni50 signal dominant (Fig. 7). But one also sees the underlying inter-decadal signal.Clearly, there was a change in the early to mid-1970s to a situation of more frequent, more intense El Niiio characteristics. Whether the period actually began withthe 1972 El Niiio or the 1976 El Niiio is problematic. The 1972 event was intense and, from a visual inspection of the time series, seems to have established a new upper level that carried through the following decade and provided an elevated base out of which the ctmonstern 1982-83 event arose. However, between the 1972 and 1976 events, the system fell back brieflyinto an opposite cool phase. After the 1976 event there were no more of what most would cal1 until after the mid-1980s. The 1972 event, while intense in the tropics (Fig. 7), did not have much effect in the northern part of the Pacific. For example, temperature time series in the Gulf of Alaska show an abrupt shift fromquite consistent negative anomalies in the first half of the 1970s to consistent positiveanomalies in the second half of the 1970s and early 1980s (Royer, 1983;Cole and McLain, 1989). However, many biological changes in regions of the world’other than the North Pacific seem to begin somewhat in synchrony with the 1972 event (e.g., the Sardinellu collapse and Balistes outbreak in the tropical eastern Atlantic). Also, if one examines the series presented by Bakun (1992) forthe North Atlantic,one finds that the apparent inflexion points most in ofthe series, including those off western Africa, seem to coincide better to 1972 (or 1973) than to 1976. Bakun (in prep.) points tothe fact that certain of the characteristics of 1972 event tended to be of greater intensity later in the year relative to other El Niiios as being perhaps at least part of the reason for its lack of effect in the highlatitude northern hemisphere. The commonly cited ENS0 teleconnection is through the Subarctic Low pressure cells (Fig. 8). Accordingly, this particular mode of teleconnection must be primarilya northern hemisphere 67
winter phenomenon; during the summer these low pressure cells tend to be nearl totally relaxed and the intense cyclogenisisthat characterizes the cooler half of the year is absent. Hisard (1988) has outlined the elements of a &-opical>>teleconnection Niiio~ between the Pacific and Atlantic Oceans. He notesthat the >and the offshore oceanic . This increased pressure gradient supports an intensified alongshore geostrophic wind. Correspondingly increased Ekman transport, directed90" to the right ( i n the northern of the wind hemisphere) leads to enhanced coastal upwellingwhich may feed back as further enhanced land-sea temperature contrast.
'. -- --.
L,
------___
..
I
76
J
.
GLOBAL CLIMATE CHANGE
f a.
b.
OPTIMAL ENVIRONMENTAL WINDOW I
I
-6ms-1
I I
MODERATE
Figure 3. (a) Diagram of a long-term upward trend descriptive of ccgreenhousen-related intensification of coastal upwelling, as wellas of associated increases in total primary organic productivity of affected marine ecosystems. (b) >gill nets. The last conclusion is that, if food transfer between the two ecosystems, the upwelling one and the oceanic one, is not achieved through systematic regular predation on Sardinella. And other small neritidpelagic fish, less abundant but more mobile could be involved, such as small Carangids. But a systematic coupling remainspossiblethrough other ways,suchexportation of macrozooplankton drifting towards oceanic ecosystem within upwelled waters. (Binet, 1991). A final paradox is that predators having advantageof Sardinella abundance increase in this upwelling ecosystem could be the dolphins (Delfinus sp Their abundance seems to have increased off CGte d’Ivoire during recent years and they are commonly observed at proximity of Sardinella schools by fishermen. Echolocation of dolphins makes themable to chasein turbid waters.
L-
%...)
REFERENCES Amon Kothias J B, F X Bard et A. Hervé, 1992.- Description des pêches et des statistiques de poissons-porte épée au port d’Abidjan, 1987-1992. SCRS/92/148 1Op Bard F X et O Pezennec 1991 - Analyse des contenus stomacaux desalbacores (T.albacares) pêchés B la senne dans le Golfe de Guinée. ICCAT Rec Doc Sci 35(2) : 1-7. Binet D. 1991 Dynamique du plancton dans les eaux côtières ouest-africaines: ecosystèmes équilibrés et déséquilibrés. In Cury et Roy Eds. Pêcheries Ouest Africaines . Variablité, instabilité et changement. Editions de I’ORSTOM ~117-136. Cayré P. and C. Roy 1986. Variabilité des rendements en albacore (T.albacares) et listao (K. pelamis) en relation avec les anomalies interanuelles de la température de surface. ICCAT Rec.Doc.Sci. 25 :67-76. Cayre P. J.B. Amon Kothias, T. Diouf et J. M. Stretta- 1988. Biologiedes thons. Chap.6 In Fonteneau et Marcille Eds. ressources, pêche et biologiedes thonidés tropicaux de l’Atlantique Centre Est.FA0 Doc Tech Peches 292 :157-268. Dragovitch A and T Potthof 1972 - Comparative studyof food of skipjack and yellowfin tunas of the Coast of West Africa. Fish BullNOAANMFS 70(4) :1087-1 1 1O. Dragovitch A. 1970 - The food of skipjack and yellowfin tunas in the Atlantic Ocean. Fish. Bull. NOAANMFS 68(3) : 445-460 Fonteneau A, T Diouf et M. Mensah 1988 - Les pêcheries thonières de l’Atlantique Tropical Est Chap 4 In Fonteneau et Marcille Ibid : 33110. Fonteneau A. 1991 - Pélagiques côtiers, pélagiques hauturiers et fluctuations environnementales: QuelquesCléments de comparaisonIn Cury et Roy Eds. Pêcheries Ouest Africaines (...) Editions de 1’ORSTOM : 31 1319 Havard Duclos F. 1972 - Synthèse Ethologique surles thons. CNEXO Rap.Sci. et Tech. No 6, 30p. Laurs R.M., Fielder P., Montgomery D.1984- Albacoretuna catch distributions relative to environmental features observed by satellite. Deep Sea Res. 31 ~1085-1099. Marcha1 E. 1959 - Analyse de quelques contenus stomacaux de Neothunnus albacora (Lowe). Bull IFAN 21(3) :1123-1136.
91
Mensah M.A. and Doyi B A 1992 The billfish fisheryin Ghana ICCAT SCRS/ 92/75. 14p ICCAT Workshop Billfish Program. Pezennec O. 1994 -Instabilité et changementsde I’écosysthme pélagique c6tier ivoiro-ghanéen: Variabilité de la ressource en Sardinelle, faits, hypothèses et théorie. Thèse UniversitéBrest, 250 p. Stretta J M, 1988 - Environnement et pCche thonière en Atlantique Tropical Oriental Chap. 7 JI Fsnteneau et Marcille (ibid).
Figure 1 : (A-Upper figure) Total catches,in metric tons,of yellowfin in Eastern Tropical Atlantic by 2cm length classes.Source : ICCAT. SCRS 1992 Report. (B-Lower figure) Length frequencies by 1 cm classes of tuna caught by Tema based baitboats, as sampled randomly for size and species 1984-1992. Figure 2 : Map of fishing grounds of Abidjan based purse seiners and Tema based baitboats. The defined area of interaction between the coastal upwelling ecosystem and the oceanic ecosystemis shown. See text. Figure 3 : Synthetic representation for years 1969-1990 of the purse seiners total catches in the area interactionas mapped in figure 2. The size of the circle is proportional to the catch. White sectors are yellowfin, striped sectors are skipjack, black sectors are bigeye.(Source Fonteneau). Figure 4 : Captures in tons of large yellowfin (category+30 Kg) for years 19801991 in the 5x5 degree squares( CWP 1 0000 and 4 0000) including the Ivorio-Ghanaian upwelling. Catches by the whole fleet of purse seiners based in Abidjan. After 1991 these complete data are not available. Figure 5 : Maps by 1xl degree squaresof the catchof skipjack and small tunas by Tema based baiboats for years 1984-1992 . Source: logbooks collected by FRUB and CRO.
92
6000
I'
1980-1991
c
Abidjan Purse seiners YELLOWFINTUNA
--d
4000
2000
J;b,,
\ \\
150
170 Lf (cm)
Senual moturity
\
\ \
4
\
O
.-::
i
r
g
I l
4 1oooor n
5
iI I
l
.......................
\
i I
\ \ \ I
BIGEYE
-
1984 1992
\
I \
TEMA BAITBOATS
\
\\
LF en centimètms
Figure 1 : (A-Upper figure) Total catches, in metric tons, of yellowfin in Eastern Tropical Atlantic by 2cm length classes. Source: ICCAT. SCRS 1992 Report. (B-Lower figure) Length frequencies by 1 cm classes of tuna caught by Tema based baitboats, as sampled randomly for size andspecies 1984-1992.
93
3
Figure 2 : Map of fishing grounds of Abidjan based purse seiners and Tema based baitboats. The defined area of interaction between the coastal upwelling ecosystem and the oceanic ecosystem is shown. See text.
94
0
..I
-2
O
5
L 1 1 . L
I
1 . 1
I
I
'1""
1Y69 1% (1 1:1:/ !
1972 1Y73 1974 197%
1976 1977 1 YYU 1Y7Y 1980
1981 1982 1983 1984 1985 1986 1987 1988 1989 1990
Figure 3 :Synthetic representationfor years 1969-1990of the purse seiners total catches in the area interaction asmapped infigure 2. The sizeof the circle is proportional to the catch. White sectors are yellowfin, striped sectors are skipjack, black sectors are bigeye.(Source Fonteneau).
95
O ID
O O
8 O (D
O O
O
W O
O
O
O
CT
:
O
O :
-'
O
E
O
b86 1 CO6 1
Figure 4 : Captures in tons of large yellowfin (category +30 Kg)for years 1980-1991 in the 5x5 degree squares( CWP 1 O000 and 4 0000) including the Ivorio-Ghanaian upwelling. Catches by the whole fleet of purse seinersbased in Abidjan. After 1991 these complete data are not available.
Figure 5 : Maps by 1x1 degree squaresof the catch of skipjack and small tunas by Tema based baiboats for years 1984-1992 . Source: logbooks collected by FRUB and CRO.
97
Dynamies and uses O
elling O€€and Ghana
Denis BINET
.B.P. 1849. P 44837 Nantes eede
25 novembre 1993
~~
Dramatic variations s f the Sardinella aurita fishery were observed since the beginning of the industrial purse seine fishery’ during the 1966s. A collapse occurred in 1973, after that a huge catch has beendone the preceding year; but against every experts opinions, the stock recovered and unprecedented sustained landings were recorded from the mid-1980s up tothe early 1990s. The fishery spred westward along theCote d’Ivoire and the fishing season almost extended al1 along the year. The fishing effort cannot account for these variations. Environmental changes were scmtinised and many papers show that they could explain a large part of the observed catch variability. Due to its thems-haline preferences, the availability of S.aurita is directly related to the upwelling intensity, and negatively to the rivers outflows.The recruitment too,is correlated to thestrength of the upwelling, through the plmkton production and thefeeding of larme. Several surplus production models could account for the landing variability, up to the early 1980s. But afterwards, the stock expansion is not explained by the former models and new hypotheses were done. A second population of S. aurita was supposed settling in the western Cote d’Ivoire, in connection to possible changes in the circulation pattern. A lengthening of the previously called could have increased the forage 98
production and hence, the recruitment issued from the first spawning period. Finally a slight cooling along the Eastern side of the Ivorian gulf could have favoured the expansion of the population from Ghana towards Cote d’Ivoire. Taken separately, none of these mechanisms can account for the whole of the observed changes, but together, they seem to explain a part of the complex stockenvironment relationships.
Résumé : Des changements surprenants de la pêcherie de Sardinella aurita ont été observées depuis le début de la pêcherie industrielle de seine coulissante dans les années 60. Un effondrement a eu lieu en 1973, après une très forte capture 1’annéeprécédente. Mais, contre tout avis des experts, le stock a récupéré et des captures record ont été régulièrement observées depuis le milieu des années 80 jusqu ’audébut des années 90. La pêcherie s’est étendue vers l’Ouest ivoirien et la période de pêche s’étend désormais tout au long de 1 ‘année.Ces changements ne sont pas attribuables à l’effort de p2che. Les changements de 1’environnement ont été soigneusement examinés et plusieurs études montrent qu’ils peuvent expliquer un bonne part de cette variabilité des captures. En raison de ses préférences thermo-halines, la disponibilité de S. aurita est directement liée à l’intensité des upwellings et inversement au débit en mer des fleuves. Le recrutement est également corrélé avec laforce del’upwelling, via la production de plancton etla nutrition des larves. Plusieurs modèles de production ont pu rendre compte de la variabilité des prises jusqu’au début des années 80. Mais depuis, 1’expansion du stock, n’est pas expliquée par les modèles précédents et de nouvelles hypothèses ont été faites. Une deuxième population de S aurita pourrait s’être développé dans 1’Ouestivoirien, en relation avec des changements dans les courants côtiers. Un renforcement de la saison marine connue comme la petite période d’upwellirtg pourrait avoir accru la production de nourriture et donc le recrutement issu de la première période de ponte. Enfin un léger refroidissement à 1 ’Est dugolfe ivoirien pourrait avoir favorisé une expansion de la population du Ghana vers la Côte d’Ivoire. Pris séparément aucun de ces mécanismes ne peut rendre comptede l’ensemble des changementobservés, mais ensemble, ils semblent expliquer pour une part les relations conzplexe entre le stock et l’environnement.
99
P. Introduction
Most coastal pelagic fisheries show variations in annual yield which are usually larger than those of demersal fisheries. However, are these variations dependent on availability or do they reflect real changesin the stock biomass? Although the real abundance of demersal fish is most accurately indicated by the catch per unit of effort (CPUE), it is difficult to assessthat of pelagic species due to problemsin estimating fishingeffort for a non-randomly distributedprey. On the one hand, the social behavior of fish, scattered or schooling, as well as their vertical migrations, make it very difficult to assess their real abundance correctly. On the other hand, the situationof fishermen searching for fish is quite different from thatof random sampling.As Mendelssohn and Cury(1 987) point out, fishing effort may introduce a bias if the variance of the catch is small compared to that of the effort,in whichcase the catch might be viewed as constant For al1 these reasons, the yearly amount ofthe catch may sometimes be the only index indicating real abundance variations.
2. History of the fishery in recent decades
l
The pelagic resources of the northern coast of the Gulf of Guinea havelong been exploited by traditional means such as dug-out canoes and beach seines. The ali net (a drifting gill net) has been use since in about1850 in Ghana (Lawson and Kwei, 1974, in Korenteng, 1991). However, there were very few reportsof variations in fish abundance before the beginning of the industrial purse seine fishery in the Côte d’Ivoire and Ghana the in 1960s. Oral traditions among tribes of fishermen might provide additional information, but enquiries have apparen not been undertaken. The causes of abundance variations can only be detenmined iE fish populations are regularly sampled in a sufficiently large fiskery. In this as in ‘ respect, al1 indications are that fishing activityhasneverbeenasgreat recent decades. With few exceptions, such as the reportof larger Sardinella aurita during the1940s in Ghanaian waters and the dominance of this species off Abidjan around 1960 (Mensah and Blanchard, pers.. comm., in Binet et al., 1991), Our i’ knowledge of variations in pelagic fish abundance is limited to the periodfrom the 1960s to now, especiallyin the absenceof sedimentological recordsof scales and other skeletal pieces. The most abundant pelagic species are Sardinella aurita, S. maderensis, Engranlis encrasicolus, Braclzydeuterns auritus and, in certain years, Seomber japonicus. During the last two decades, the most striking events werethe surge in triggerfish (Balistes carolinensis) population during the 1 9 7 0 ~followed ~ recently by a sharp decline (Caverivikre, 1991, 1993),and the dramatic variations in S. aurita landings. The two Sardinella species constitute most ofthe pelagic fishery. Until 1980, most of the catch from Côte d’Ivoire purse seiners consisted of Sardinella maderensis. Since then, S. aurita has ranked first in Côte d’Ivoire landings. 100
The history of the S. aurita fishery during the last decades may be divided into three periods (Fig. 1) :
(i) During the 1960s, landingsin the Côte d’Ivoire and Ghana were moderate but variable, ranging between 8,000 and 47,000 mt. An approximate assessment of fishing mortality (Marchal i n FAO, 1974) determined thata sustained catch could not exceed 30,000 mt; butal1 landings by traditional fishermen and baitfish caught by tuna fishing boats were notincluded. (ii) In 1972, the catch suddenly increased to 95,000 mt, almost entirelyfished by canoes in Ghanaian waters. After this unusually highcatch, S. aurita virtually disappeared. After a 5-year collapse, the species progressively increased again and had completely recoveredby 1978. (iii) Beginning in 1981,a dramatic expansion took place, with highcatches over several consecutive years. From 1985 to 1987, the stock yielded successively 106, 90 and 82 thousand mt per year withno collapse after such large harvests. Population dynamics cannot clearly account for the small variationsin yield during the 1960s, the highcatch of 1972 or the sizeable changes occurring later. The role of the environment in S. aurita availability was soon noted (Marchal in FAO, 1974) and put forward to explain the 1972 overfishing. Annual recruitment later on was also related to environmental variations. Wise management of living resources requires an understanding of climatic forcings on the ecosystem in order todetermine the appropriate level of fishing effort. The purpose of this paperis to analyzethe different hypotheses proposed in an attempt to explain how the various sources of fluctuations have affected this fishery.
2.1. Environmental variables and availability: River floods and upwelling Although the occurrence of S. aurita in coastal surface waters was first related to cold seasons, it soon became apparentthat this species, in an area where the rainy season and the long cold season roughlycoincide, was influencedmore by high salinity than cold waters (Marchal in FAO, 1974). During previous decades (the 1960s and 1 9 7 0 ~S. ) ~aurita avoided waters inthe western part of the Côte d’Ivoire shelf where salinity was lessened by rains and riverfloods, so that most of the catches came from the eastern side. During the first months of the year, the highest salinities (>35 per mil) occur the central in area where the best catches are made. These seasonal and geographicalrelationships between high salinity and the occurrence of S aurita are also apparent when yearly catches are compared with upwelling intensity index variations 2a) (Fig. (Binet, 1982). However, several years are clearly out of the correlation range, namely 1972, the overfishing year, and 1973-1977, the recovery period. Infact, why was 1972 so distant from the correlation straight line? 101
2.2. The oncoming drsugkt Before the 1973 collapse, a negative relationship was observed between rainfall levels and the amount of S. aurita landed in Abidjan duringJuly-August (Marcha1i n FAO, 1974).Bakuns (1978) noted the same relationship for the Ghana fishery from 1964 to 1968, with periods of extremely good fishing being associated with strong upwelling and subnormal coastal rainfall.The negative relationship between upwelling and pluviornetry was observed alongal1 eastern boundary currents. The occurrence of upwelled waters cools the lower atmospheric layers andprevents the development of vertical cloudswhich cause rain. This geographical relationship has also been reported among time series indicating that interannual variability of upwelling and coastal pluviometry are correlated (Hisard, 1986; Binet, 1982). The first year of severedrought in subsaharian Africa was 1972. The input of fresh water into the Gulf of Guinea was further reduced by the building of two dams across the Volta and Bandama rivers, and the salinity of coastal waters was not lowered by terrestrial runoff. That year, a large numberof young S. azwitn were harvested nearthe Coast. Binet (1982) conjectured thata close approach to the shore, usually prevented by their halophilous behavior, occurred in 1972, leading to recmitment overfishing. In Senegalese waters, that year was also marked by high catchesof S. nurifa,the levelof which has never beenequalled (Fr'rCon, 1991a). A negative relationship was found between annual catches (1963-1972 and outflow of the 1978-1979) of S. nurita from the C6te d'Ivoire and Ghana and the Bandama and Como6 rivers (Pig.2b). The 1973-1977 points, during which the stock was recovering, were aligned along a different straight line. These two results allowed computation ofa regression between catches, upwelling and river flow, showing the climatic influence on availability (Binet, 1982).
'
,
,
2.3. An inaccessible part of the biomass L
These results led Laloe (1 988) to suppose thata part of the stock is ordinarily unavailable to the fishery and then to developa production mode1 based on the assumption ofan inaccessible part of the biomass. In the case of a surplus production model, itis generally supposed that biomass production is a function of exploited biomass. However, if a certain quantity of the biomass is out of reach of the fishery, this is not correct. Laloe proposed two modelsbased on the assumption that a part of the stock biomass is not available except under a riverflow threshold at which the whole biomass becomes available. Laloe started from the Graham-Schaeffer model: dB/dt = B,.H.(B,-BV)- q.f,.B, (1) which describes the instantaneous change of the biomass as a function of exploited biomass(BJ, unexploited biomass(BJ, fishing effort(fJ at time t, and catchability (q), H being the growth rate of the stock.
1 O2
I
He assumed that the fishery only acts on a part of the stock, depending on the available biomass, a being the rateof inaccessibility: dB/dt = B,.H.(B,-BJ - q.f,.(Bïa.BV) (2) The value of a is constant except during very low river flow (1972, 1978) when it is equal to O. Al1 the biolnass then becomesavailable, giving equation (1) (Fig. 3b). Another constraint is that q=O if B, < a.Bv. This model describes catch series correctly uhtil 1980. Most interestingly, it best fits observed data during the ccrecovery period>>of 1973-1976 (Fig 3a).
3. Upwelling, availability and recruitment The following models include the short or mid-term hypothesized effects of upwelling on CPUE (Le., respectively availability and recruitment). Unfortunately, they take either the whole pelagic fishery into consideration (Cury and Roy, 1987; Mendelssohn and Cury, 1987) or the onlySardinella maderensis fishery (Mendelssohn and Cury, 1989), but notS. aurita alone. The theoretical basis for such straightforwardnessin modeling is that an overexploited species may be replacedby another, so that the biomass of a system is more steady than that of its components. Moreover, this plurispecific approachdoes not account for the diversity of ecological niches or the difference in species life-spans.
3.1. Use of a generalized stock production model In considering the unexploited biomassBv as a function of a climatic factor (CLIM), B,,avg beingthe average biomass and b aconstant, Bv = Bvavg(1 + b.CLIM), Cury and Roy (1987) obtainedthe following model which Fréon (1984) had studied earlier: C p m , = e-Ao.li .(A + B CLIM,) where'CPUE,is the catch per unitof effort of the yeari, fi fishing effort; and Ao, A, B the constants. Cold waters reaching the surface bring nutrientsinto the euphotic layer, so that the planktonic food web is boosted by these inputs. It is considered that strong upwelling can enhance the stock biomass by accelerating individual growth or decreasing the natural mortality of these plankton-feeders (ORSTOM,1976), although theexact biological mechanisms involvedare unknown. However, the abundance of Sardimlla larvae is related to upwelling intensity, and recruitment is completed at one year.Thus, the following equation gives an empirical model, including fishing effort and an upwelling index (UPW),for years i and i-1: CPUE, = e-AD.fi.(A + B.UPW, + C.uPW,-I) This model allows testing of the roles played by the different variables (Fig. 4). Fishing effort alone can account for only 18% of CPUE variability, whereas the upwelling indices for years i and i-1 account for 40%. The best fit (73%) was obtained with the 3 variables, i.e., fishing effort and the two upwelling 103
indices. An interesting result is that upwelling intensity from the catch year (i) accounts for more vakability than upwelling during the spawning year (i-1). Does this imply that planktonic abundance,providing forage for juvenile or adult fish, is more important for the avaiilability of the resource thanfor spawning suceess and larval survivall? If so, this would mean th& availability is more important than recruitment for @PUE, providedthat no recruitment overfishing oecurred in the preceding years. 3.2. Short-term variations and A
Mendelssohn and Cury (1987) usedmultivariate auto-regressive movingA) to model C8te d’Ivoire pelagic fishCPUE as a function of sea sufice temperature (SST). Fortnightly SST and @PUEmeans were calculated for the main fishing areas, and the model then fills in any missing point. The results closely reproduce observedsekes. Thus, a cross-correlation betweenSSTand CPUE-computed seriesshows that the best fit is obtained for a one-fortnight time lag. This is approximately the time lag needed by copepods to benefit from upwelling enrichment (Binet, 1976). The exact duration of the production transfer from nutrient enrichment to the molting ofcopepodites, including diatom blowup, maturationand spawning of copepods, is usually longer. However, ontogenet vertical migration between the Guinea Current and Undercurrent c m shorten that time span (Binet, 1991).Thus, if it is detemined that short-term availability of pelagic resources is indicated by upwelling intensity within a 2-week time lag, the model is coherent with planktonic production mechanisms. Strong CPUE persistenceis mother noticeable result. Once fishare present, come to the surface and school they tendt s remain. The pelagic species probably more often when there is abundant zooplmkton on which they can feed.
gration patterns 1
Mendelssohn and Cury (1989) used space-timeanalysis to study a possible relationship between migration and hydrologic factors in Sardinella maderensis @PUEoff the Côte d’Ivoire. An east-West migration !vas apparent in @PUEfor one- and two- fortnight lags, and a reverse movementwas observed in one- and four-fortnight lags. Similar patterns were notedfor transformed SST. Favorable conditions were associated with frontal structure. The west-to-east movement probably reflects advection due to the Guinea Current, and the east-to-West movement changes in upwelling. Fisk evidentlymigrate to be in the area most favorable for zooplankton enrichment.
3.4. An exception ts the >indices. Recruitment and replenishmentof the stock biomass arealso related toplankton Thus, the same upwelling indexrnay serve as a abundance in the preceding year. proxy for al1 these parameters, providedthat it is used with differenttime lags. Nevertheless, the exceptional catch of the mid-1980s disrupted al1 correlations, so that these models proved tobe analytic but not predictivetools. During the 1980s, mpwelling>>indices showed no particular trends, but the time of cooling events and the place where they occurred showed slightdifferences compared to previousdecades. This new patternmay have favored a westward expansion of the Ghanaian population as well as boosting the previouslysmall population nearCape Palmas. Possible changes in the current pattern would also have ledto similar results. Finally, in this complexecosystem in which enrichment events did result not from a single physical factor, it is likely that each of the above-cited works provides part of the truth. However,river flow and cooling intensity,identified as the leading factors duringthe early decadesof the fishery, no longer have the same importance inthe 1980s due to large-scale changesin the environment. On a long-term basis, river runoff has considerably decreased, and surface water cooling events have increased, whereas the surronnding waters haveundergone a slow warming and oceaniccirculation has changedfor some time. Perhaps, it would be usefulto have indicesof the spatial variabilityof the two coldseasons and s' changes in the circulation pattern before introducing these factors ints models. But the main problemwe are facing now is to ltnow how long willlast these > environmental characteristics ?
114
-
REFERENCES Aboussouan A. (1971). Oeufs et larves de téléostéens de l’ouest africain. IX. Sur la migration des larves de SardinelZa aurita Val. à proximité de la presqu’île du Cap-Vert.Bulletin de Z’IFAN, 33(A) : 227-236. Arfi R., Pezennec O., Cissoko S., Mensah M.A. (1991). Variations spatiale et temporelle de la résurgence ivoiro-ghanéenne. in: Pêcheries ouestafricaines. Variabilité, instabilité et changement, P. Cury et C. Roy eds. ORSTOM, Paris, 162-172. Bakun A. (1978). Guinea current upwelling. Nature, 271, 147-150. Bakun A. (1990). Global climate change and intensification of coastal ocean upwelling. Science, 247, 198-201. Bakun A. (1992). Global greenhouse effects, multi-decadal wind trends, and potential impacts on coastal pelagic fish populations. ZCES mar. Sci. Symp., 195,316-325. avec le Binet D. (1976). Biovolumes et poids secs zooplanctoniques en relation milieu pélagique au dessus du plateau ivoirien. Cah. ORSTOM. sér. Océanogr., 14,301-326. Binet D. (1982). Influence des variations climatiques sur la pêcherie des Sardinella aurita ivoiro-ghanéennes : relation sécheresse - surpêche. Oceanologica Acta, 5,443-452. Binet D. (1991). Dynamique du plancton danseaux les côtières ouest africaines : écosystèmes équilibréset déséquilibrés. in:Pêcheries ouest-africaines. Variabilité, instabilité et changement,P. Curyet C. Roy eds. ORSTOM, Paris, 117-136. Binet D., Marchal E., Pezennec O. (1991). Sardinella aurita de Côte d’Ivoire et du Ghana : fluctuations halieutiques et changements climatiques. in: PiMeries ouest-africaines. Variabilité, instabilité et changement, P. Cury et C. Roy eds. ORSTOM,Paris, 320-342. Binet D., Marchal E. (1992). Le développement d’une nouvelle population de sardinelles devant la Côte d’Ivoire a-t-il été induit par un changement de circulation ? Ann. Inst. Oceanogr. 68 (1-2), 179-192. Binet D., Marchal E. (1993). The large marine ecosystem of shelf areas in the gulf of Guinea: long-term variability induced byclimatic changes. in: Large marine ecosystems :stress, mitigation and sustainability, Ed. by K. Sherman, L.W. Alexander and B.D. Gold. AAAS Press, 104-118. 115
Binet D., Servain 9. (1 993). Have the recent hydrological changes in the Northern Gulf of Guinea induced theSardinella aurita outburst ? Oceanologica Acta l6?247-260. Caverivikre A. (1991). L’explosion dimographique du baliste (Balistes carolinensis) en Afrique de l’ouest et son évolution en relation avec les tendances climatiques. in: Pêcheries ouest-africaines. Variabilitk, instabilid et clzangernent, P. Cury et C. Roy eds. ORSTOM, Paris, 354-367. Caverivière A. (1993). Les ressources en poissons démersaux etleur exploitation. in :Environnementet ressouces aquatiques deCôte d’Ivoire I. Le milieu marin, P. Le Eoeuff, E. Marchal, J.B. Amon Kothias eds. ORSTOM, Paris, 427-488. Citeau J., Finaud L., Cammas J.-P. and DemarcqH. (1989). Questions relative to ITCZ migrations over the tropical Atlantic ocean, sea surface temperature and Senegal river runoff.Meteorol. Atmos. Phys. 41 181190. ~
Colin C.(1 988). Coastal upwellingevents in front of the IvoryCoast during the FOCAL program. OceeansfogicaActa, 1 1 , 125-138. Colin C. (1991). Sur les upwellings équatorial etcôtier (5”N) dans le Golfe de Guinée. Oceanologica Acta, 14 (3), 223-240. Colin C., Gallardo Y., Chuchla R., Cissoko S . (1 993). Environnements climatique et ochographique sur le plateau continental de C8te d’Ivoire. in : Environnement et ressouces aquatiques de G î t e d’Ivoire 1. Le milieu marin, P. Le Loeuff, E. Marchal, J.B. Amon Kothias eds. ORSTOM, Paris, 75-1 10. Cury P., Roy C. (1987). Upwelling et pêche des espkces pélagiques côtikres de Côte d’Ivoire : une approche globale.OceanologicaActa, 10,347-357.
Cury P., Roy C. (1989). Optimal environmental window and pelagic fish recruitment success in upwelling area.Canadian Jownal of Fisheries , and Aquatic Sciences, 46, 670-680. Ecoutin J.M., Delaunay K.? Konan J. (1 993). Les pêchesartisanales maritimes. in :Environnementet ressouces aquatiques deCôte d’Ivoire I. Le milieu marin, P. Le Loeuff, E. Marchal, J.B. Amon Kothias eds. ORSTOM, Paris, 537-550. FAO, (1974). Etude et mise en valeur des ressources en poissons pélagiques côtiers. Côte d’Ivoire. Conclusionet recommandation du projet. PNUDFAO, FI : DP/IVC/66/506, Rome, 72 p. dactylogr. 116
Fréon P. (1 984). Des modèlesde production appliquésà des fractions de stocks dépendantes de vent d’upwelling (pêche sardinière au Sénégal). Océanogr. trop., 19, 67-94. Fréon P. (1988). Réponses et adaptations des stocksde clupéidés d’Afrique de l’Ouest à la variabilité du milieu et de l’exploitation. Analyse et réflexions à partir de l’exemple du Sénégal. Etudes et Thèses. ORSTOM, Paris, 287 p. Fréon P. (1991). L’introduction d’une variable climatique dans les modèles globauxdeproduction. in: Pêcheries ouest-africaines.Variabilité, instabilité et changement,P. Cury et C. Roy eds. ORSTOM, Paris, 395-424. Herbland A., Marcha1 E. (1991). Variations locales de I’upwelling, répartition et abondance des sardinelles en Côte d’Ivoire. in: Pêcheries ouestafricaines. Variabilité, instabilité et changement, P. Curyet C. Roy eds. ORSTOM, Paris, 343-353. Hisard P. (1980). Observation de réponses de type dans l’Atlantique tropical oriental, Golfe de Guinée,Oceanologica Acta, 3, 69-78. Hisard P. (1988). El Niiio response of the tropical Atlantic ocean during the 1984 year.in : /nt. Symp. Long Term Changes Mar. Fish Pop. 1986,T. Wyatt and M.G.Larraneta eds., Vigo, 273-290. Hisard P., HéninC., Houghton R., PitonB., Rua1 P. (1986). Oceanic conditions in the tropical Atlantic during 1983 and 1984. Nature, 322, 243-245. Houghton R. W. (1976). Circulation and hydrographic structure over the Ghana continental shelf during the 1974 upwelling. J. Phys. Oceanogr., 6,910-924. Ingham M. C. (1970). Coastal upwelling in the northwestern gulf of Guinea. Bull. Mar. Sci., 20, 1-34. Katz E.J., Hisard P., Verstraete J.M., Garzoli S. (1986). Annual change of sea surface slope along the Equator of the Atlantic Oceanin 1983 and 1984. Nature, 322,245-247. Koranteng K., A. (1991). Some aspects of the sardinella fishery in Ghana. in : Pêcheries ouest-africaines. Variabilitt!, instabilité et changement, P. Cury et C. Roy eds. ORSTOM, Paris, 269-277. Laloe F., (1988). Un modèle global avecquantité de biomasse inaccessibleliée : application aux données de la pêche aux conditions environnementales ivoiro-ghanéenne de Sardiaella aurita. Aquatic Living Ressources 1, 289-298. 117
Lamb P.,(1 978). Case studies of the tropical Atlantic surface circulations pattern during recent subsaharian weatheranomalies in 1967 and 1968.Mon. Weather Rev. 1 06- 462-49 1 . Lemasssn E., Webert J.P. (1973). Les courants marins dans le golfe ivoirien. Calz. ORSTOM, sér. Oe&amgr., 1 1 67-96. .1
Mahé G., (1993). Lesécsulements fluviaux sur la faGade atlantique de l’Afrique. Etude des Clémentsdubilan hydrique et variabilitC interannuelle, . analyse de situations hydroclimatiques moyenneset extrgmes. Etudes et ThAses. ORSTOM, Paris, 438 p. Marchal E. (1991). Un essai de caractkrisation des populations de poissons pélagiques côtiers: cas de Sardinelln aurita des côtes ouest-africaines. in ; P2cheries ouest-africaines. Variabilité, instabilitb et changement, P. Cury et C. Roy eds. ORSTOM, Paris, 192-200.
Marchal E., (1 993). Biologie et 6cologie des poissons pélagiques côtiers du lit ivoirien. in :Environnenzentet ressouees aquatiques de Cdte d’Ivoire I. Le milieu nznrin, P. Le Loeuff, E. Marchal, J.B. Amon Kothias eds. ORSTOM, Paris, 237-270. Marchal E., Picaut J. (1977). Répartition et abondance évaluées par échointigration des poissons du plateau ivoiro-ghanCen en relation avec les upwellings locaux. J. Rech. Océannogr., 2,4; 39-57. Mendelssohn R, Cury P. ( 1 987). Forecasting a fortnightly abundance index of the Ivorian coastal pelagic species and associated environmental conditions. “an. J. Fish. A p a t . Sci., 44,408-421. Mendelssohn R, Cury P. (1989). Temporal and spatial dynamics o f a coastal pelagic species,Sardinella mderensis off the Ivory Coast.Can. J. Fish. A ~ L MSei., z ~ . 46, 1686-1697. of climatic changeson the coastal oceanography Mensah M. (1 991). The influence ofGhana. in : Pgcheries ouest-africaines. Variabilitk, instnbilite” et changement?P. Cury et C. Roy eds. ORSTOM, Paris, 67-79. Moore D.W., HisardP., McCreary J.P., Merle J., O’Brien J.J., Picaut J., Verstrate J.M., Wunsch C. (1 978). Equatorial adjustment in the eastern Atlantic. Geophys. R ~ sLett., . 5,637-646. Morlière A. (1970). Les saisons marines devant Abidjan. Doc. Scient. C.R.O. Abidjan, 3 (2)’ 1-15.
118
‘
Morlière A., Rebert J.P. ( 1 972). Etude hydrologique du plateau continental ivoirien. DOC.Scient. C.R.O. Abidjan, 3, 1-30. ORSTOM, (1 976). Rapport du groupede travail sur la sardinelle des côtes ivoiroghanéennes. Fishery Research Unit Tema, Centre de Recherches Océanographiques Abidjan, ORSTOM.62 p. Pezennec O., Bard F.-X. (1992). Importance ecologique de la petite saison d’upwelling ivoiro-ghanéenne et changements dans la pêcherie de Sardinella aurita. Aquat.Living Resour., 5,249-259. Pezennec O., Marchal E., Bard F.-X. (1993). La pêche des petites espèces pélagiques en Côte d’Ivoire in :Environnement et ressouces aquatiques de C6ted’lvoire 1. Le milieu marin, P. Le Loeuff, E. Marchal, J.B. Amon Kothias eds. ORSTOM, Paris, 387-426. Philander S.G.H. (1986). Unusual conditions in the tropical Atlantic ocean in 1984. Nature, 322,236-238. Picaut J. (1983). Propagation of the seasonal upwellingin the eastern equatorial Atlantic. J. Phys. Oceanogr., 13, 18-37. Piton B., Wacongne S. (1985). Unusual amountsof very saline subsurface water in the eastern Gulf of Guinea in May 1984.Tropical Ocean-Atmosplzere Newsletter, 32, 5-8. Roy C. ( 1 992). Réponses des stocks de poissons pélagiquesà la dynamique des upwellings en Afrique de l’Ouest : analyse et modélisation. Etudes et Thèses. ORSTOM, Paris. 146 p. Shannon L.V., Boyd A.J., Brundrit G.B., Taunton-Clark J. (1986). On the existence of an El-Nifio-type phenomenon in the Benguela system. J. mar. Res., 44,495-520. Sinclair M. (1988). Marine populations: An essay on populationregulation and speciation. Washington Press,Seattle and London.252 p. Verstraete J.-M. (1992). The seasonal upwellingsin the Gulf of Guinea. Prog. Oceanog. 29, 1-60.
119
Figure 1 : Annual catches of Sardinella aurita in Ghana andCôte d'Ivoire from 1963 to 1990, according to Peaennec and Bard, (1992). ure 2 : Relationships between the annual yield (Y) of Snrclinella aurita in Ghana andCôte d'Ivoire (1964-1985) and (a) : an index of upwelling intensity (U) at Tema, (b) : the Bandama and Corn06 riversflows The regression are (a) Y(t) = 162U - 5430 ( ~ 0 . 4 4 and ) (b) Y(t) = 50.3 D(m3/s) -t 56700, (r=0.45). From Binetet al., (1991).
(m.
Figure 3 : Global production model with unaccessible part of biomass (Laloe, with model M l . b) Equilibrium 1988). a) Observed and fitted catches, Catch-Effort relationship when biomass is totally accessible (a=@ or not (a=0.453), with model M2. Figure 4 : A production model using fishing effort and upwellingindices (Cury and Roy, 1987). @PUEfor al1 the pelagicspecies, from 1966 to 1981, observed values (solid line) and predicted values (dashed line), with the following parameters : 1) effort only; 2) upwelling indices, year i and i-1only; 3) effort and upwelling index, year 4) effort i; and upwelling 1 ;5) effortand upwelling indices, yeari and i- 1 . index, year i-
Figure 5 : Production model using fishing effort and environmental indices (Fréon, 1991). (a) : Observed and modeled@PUEof the Ivorian purse seiners using upwelling indices for abundance and catchability of Sardinella nurita. Adjustement withoutthe years 1974 and 1975. (b): Observed and modeled S. ar4rifa catch of the Ivorian and Ghanaian fisheries using upwelling indexand river flow to aceount for abundance : recovering period following and catchability, respectively. 1974-1 978 the 1972 oveffishing. Figure 6 :Density of pelagic fish estimated by acoustic surveys(thille?) over the Ghana and Côte d'Ivoire shelf in 1976, 77, 80, 81 86 and 87. For 1981, the average for two cruises is given; Balistes cnrolinensis amounted 25-83 % of this biomass. During subsequent surveys Sarclitzella aurifa represented 10 to 38 %, From Marcha1 (1993) and Binet and Marcha1( 1 992).
Figure 7 : Annual flows (1955-1 988)in m3/sfor the Bandama,Comoé and Volta rivers. Dataof Bandama, ComoC and Volta atSenchi from BRSTOM and DRES Côte d'Ivoire, data for the Volta river at Akosombo from the VoltaRiver Authority (Arma, pers. comm.). Figure 8 :Map of the northern part of the Gulf of Guinea, indicatingthe positions of coastal stations along the Ivorian gulf and the 2" x 2' (S1 to S 5 ) grids used inthe processing of ships observations.
.
120
Figure 9 : Annual anomalies of : (a) wind stress modulus m2/s2; and (b) SST (OC) averaged betweenO" and 10"W(S 1 to S5)from ships data (1 9641990). Dashed lines represent adjusting functions. From Binet and Servain (1993). Figure 10 : Annual differences (1970-1990) between SST in January-March (little cold season) and in July-September (great cold season), from the merchant ship data setin the Ivorian region (8"-2" West, 4"-6"N). From Pezennec and Bard (1992). Figure 11 : Annual differences (1964-1990) between eastern and western sea surface temperature in the Ivorian gulf : SST (4"-2"W) - SST (8"6"W), for January-March and July-September. The usual SST zonal gradient is inverted during the great cold seasons 1981- 1 985 and during the little cold seasons 1986-1989. From Binet and Servain (1993). Figure 12 : Outlines of the changes in the circulation of the tropical Atlantic, due tothe Inter-Tropical Convergence Zone (ITCZ) move towards the equator in 1984. NECC : North Equatorial Counter Current, GC : Guinea Current, GUC : Guinea Under Current. Italics : supposed changes. From Binet and Marcha1 (1 992). Figure 13 : Longitudinal section of the zonal componentsof the circulation off the Côte d'Ivoire to the Nigeria; 0-100 m profiles on the continental shelf at 10 to 23 nautical miles from the shore, May 1972. The zonal velocities are indicated in cm/s; eachcontour line corresponds to a 1O cm/s variation. E and W indicate the eastward and westward flow, respectively. The O-contour is the shear layer separating the Guinea Current (GC) fromthe Guinea Under Current (GUC). From Lemasson and Rebert, (1 973). Figure 14 : Annual changes (1 971-1 990)of the mean ITCZ latitude in FebruaryApril, in1/1O degrees latitudeN. Data from remote sensing observations (UTIS-Dakar and Citeau pers.comm.). Figure 15 : Regression coefficients of SST vs Alongshore Wind Stress, for every month andsquares 1 to 5. 12th differencing each monthly series has been done prior to analysis to remove autocorrelations. Negative values indicate likely wind-induced upwellings. The interannual variations of the surface waters coolings in the long cold season are unrelated to the wind stress, meanwhile weak upwelling trends occur during the rest of the year, except in SI. Longitude ranges of the geographical squares S1 : 1 0"-8"W, S2 : 8"-6"W, S3 : 6"-4"W, S4 : 4"2"W, S5 : 2"W-O" (See Figure 8). From Binet and Servain, (1 993).
121
Sardinella aurita catches (t) 90000 80000 70000 60000 50006 40000 30000 20000 10000 0 M
6
M
6
z
en P
en
eo
CD
a, e
P
-0-
GHANA
aQ
m e
COTE D'IVOIRE
Figure 1 : Annual catches of Surdinellu auritu in Ghana and Côte d'Ivoire feom 1963 to 1990, accoeding to Pezennec and Bard, (1992).
122
Fig. 2 a
l
65
"1
ltl
PU
67
401
82
20 10
7;
71
z,
,,
7
64
73
,
,
166
1
75 74
O
Fig. 2 b
1101
85
I 0 100-
72
,O
86
a 90-
U
2 8070E 6 60.- 50r i 40-
al ..-=
83 84 67
70
50
69
88' 66
79
O.
82 64
71
302010-
78
7365
60 I
1
150
1
76
77
63 74 1
250
75 I
1
1
+
350, ,450 Indice d upwelling a Tema
Figure 2 :Relationships between the annual yield (Y) ofSardirlellualcritainGhanaand Côted'Ivoire (1 964-1 985) and (a) :an index of upwelling intensity (U) at Tema, (b): the Bandama and Comoé rivers flows (FI. The regression are (a) Y(t) = 162U - 5430 (r=0.44) and (b) Y(t) = - 50.3 D(m3/s) + 56700, (-0.45). From Binet et al., (1991).
123
Figure 3 : Global production model with unaccessible part of biomass (Laloe, 1988). a) Observed and fitted catches, with model Ml. b) Equilibrium Catch-Effort relationship when biomass is totally accessible (a=O) or not (a=0.453), with model M2.
124
15.1
PUE
@
0,- , , , , , , , 66 70
,
15-
, , , ,
75
, , ,.
m ANNEES
PUE
O
...
O , , . , , , . . , , , . . 6 6 7 0
75
81 ANNEES
Figure 4 : A production mode1 using fishing effort and upwelling indices (Cury and Roy, 1987). CPUE for al1 the pelagic species, from 1966 to 1981, observed values (solid line) and predictedvalues (dashed line), with the followingparameters : 1) effortonly; 2) upwelling indices, year i and i-1 only; 3) effort and upwelling index, year i; 4)effort and upwelling index, year i-1; 5) effort and upwelling indices, year i and i-1 .
125
66
76
75
86 Annies
Fig. 5 b
66
70
75
88
Ann$@s
Figure 5 : Production mode1 using fishing effort and environmental indices (Frêon, 1991). (a) : Observed and modeled CPUE of the Ivorian purse seiners using upwelling indices for abundance and catchability of Sardinella aurita. Adjustement without the years 1974 and 1975. (b) : Observed and modeled S. auritu catch of the Ivorian and Ghanaian fisheries using upwelling index and river flow to account for abundance and catchability, respectively. 1974-1978 : recovering period following the 1972 overfishing.
126
Biomass obtained by acoustic surveys tonnes l nilled 70
50
50
10
30
20
10
O
Figure 6 : Density of pelagic fïsh estimated by acoustic surveys (t/milleL)over the Ghana and Côte d'Ivoireshelfin1976,77,80,81,86and87.For1981,theaveragefortwocruisesisgiven; Balistes carolinensis amounted 25-83 % of this biomass. During subsequent surveys Sardirzellu auritu represented 10 to 38 %, From Marchal (1 993) and Binet and Marchal (1992).
-
-
VOLTA BANDAMA COMOE
I -.-
Senchi -Akosombo
-*-
BandaCom
I
Figure 7-: Annual flows (1955-1988)in mYs for the Bandama, Comoé and Volta rivers. Data of Bandama, Comoé and Volta at Senchi from ORSTOM and DRES Côte d'Ivoire, data for the Volta river at Akosombo from the Volta River Authority (Arma, pers. comm.).
127
L
128
b
-
a Wind Stress Modulus Anomaly(mA/sA) O" 10" w
-
-1 -2
-3 -4 -5 m --
-6
Sériel
-
-
Série2
-
b SST Anomaly (OC) 0"-10"W 1
03 096
0,4
072
O -0,2 -0,4 -0,6
-0,8
Sériel
-
Série2
I
Figure 9 : Annual anomalies of: (a) wind stress modulus mYs2; and (b) SST ("C)averaged between O" and 1O"W (S1 to S5) from ships data (1964-1990). Dashed lines represent adjusting functions. From Binet and Servain (1993).
129
a
ANNEES
Figure %O :Annual differences (1970-1990) between SST in January-March (little cold season) and in July-September (great cold season), from the merchant ship data set in the Ivorian region (8"-2" West, 4"-6"N). From Pezennec and Bard (1 992).
-
-
East West [ship] SST differenees (S4 $2)
0,50
0,oo -0,50
-1 ,O0 -1,50
-2,oo
m--
January-March
-
July-September
Figure 11 : Annual differences (1 964- 1990) between eastern and western sea surface temperature in the Ivorian gulf: SST(4"-2"W)- SST (8"-6"W), for January-March and July-September. The usual SST zona1 gradient is inverted during the great cold seasons 1981-1985 and during the little cold seasons 1986-1989. From Binet and Servain (1993).
130
131
3"W
Abidjan
3"E
6" I
I
Takoradi
l
Cotonou
Aecm
Lagos
Figure 13 : Longitudinal section of the zona1 components of the circulation off the Côte d'Ivoireto the Nigeria; 0-100 m profiles on the continental shelf at 1O to 23 nautical miles from the shore, May 1972.Thezonal velocities areindicatedincds; each contourline corresponds to a 10 c d s variation. E and W indicate the eastward and westward flow, respectively. The O-contour is the shear layer separating the Guinea Cunent (GC) from the Guinea Under Current (GUC). From Lemasson and Rebert, (1973).
-
latitude February
1983 1980 1977 1974
-
1986
1989
-10,oo
Figure 14 : Annual changes (1971-1990) of the mean ITCZ latitude in February-April, in 1/10 degrees 1atitudeN.Datafrom remotesensingobservations (UTIS-DakarandCiteau pers. comm.).
132
Slopes of the regressions SST vs AWST 03 0,6 04 02 O
-0,2
-0,4
-0,6
-0,8
Figure 15 : Regression coefficients of SST vs Alongshore Wind Stress, for every month and squares 1 to 5. 12th differencing each monthly series has been done prior to analysis to remove autocorrelations. Negativevaluesindicatelikely wind-inducedupwellings.Theinterannua1 variations of the surface waters coolings in the long cold season are unrelated to thewind stress, meanwhile weak upwelling trends occur during the rest of the year, except inS 1. Longitude ranges of the geographical squares S1 : 10"-8"W, S2 : 8"-6"W, S3 : 6"-4"W, S4 : 4"-2"W, S5 : 2"W-O" (See Figure 8). From Binet and Servain, (1993).
133
EY '
Abstraet : (In artisanal marine fisheries on the Ga-Dangme Coast) The paper discusses social aspects of labour arrangements in canoe fisheries among the Ga-Dangmes, with particular reference to the place of kinship in labour contracts. Important social features of labour relation inthese fishing communities include : the dependence on the family as a labour pool ; family-based informal training and financing arrangements ;pressure on 'owners of capital assets' (canoe owners, fish smolcers/traders) to admit family members into their enterprises, resulting in overcrowding ; a local avelfare/socialsecurity network that encourages ' their an extensive redistribution of wealth fronl 'owners 'of capital a ~ s e f sto 'worker' relatives ; the incorporation of authority structures in fishing into local eommunity political pswer structures. The implications of this overlap between social interests and business interests for the future of the Ga-Dangrne artisanal fisheries industry are addressed. 111 conclusion it is noted that kinship ties promote effective labour management. On the other hand they frustrate the bargaining position of both canoe owners and their crew, and may thwart innovation in labour transactions.
Ce document discute desaspects sociaux de 1'organisation du travail dans la pgcherie pirogrtibre chez les Ga-Dangmes. Le r6le de lcl parentde dans ces relations est particlllibrement traite?.Les caracte'ristiques importantes dutravail familial dans cesconzmunautés de gzcheurs incluent :
1. Dr. Ellen Bortei-Doku Aryeetey is a Research Fellow (Sociology) at the Institute of Statistical, Social and Economic Kesearch (ISSER), University of Ghana, Legon.
134
La dépendance de lafamille comme resewe de main d’oeuvre; la formation et les arrangements financiers dans la famille ;la pression su.r les possédants des investissements effectués (dans lespirogues, le &mage et les commerce du poisson) afin qu’ils engagent des membres de la famille dans leur en.treprise,ce qui entraine une surcapacité de main d’oeuvre ; U n système de protection sociale qui encourage la redistribution extensive des richesses acquises par les investisseurs vers les parents des travailleurs ; L’incorporation des structures hiérarchique des pêcheurs dans les structure de pouvoir politique des communautés locales. Les implications de ce recouvrement entre les intérets sociaux et les intérets des en.treprises,pour lefutur des pêcheries artisanales Ga-Dangme sont traitées. En conclusion il est établi queles liens de parentèle poussent a un gestion effective de la main d’oeuvre. En revanche cela affaiblit les capacités de négociation à la fois des armateurs des pirogues et de leurs iquipages, ce qui peut stériliser l’innovation dans l’établissement de relations de travail.
1. Introduction Ghana continues to be heavily dependent on artisanal marine fisheries for over 70 percent of the fish landedlocally, an impressive increasefrom about 40 percent of the total national catchin the late 1960s (Fisheries Research Institute, 1994; Fisheries Division, 1966-67). This industry is therefore an important source of relatively cheap animal protein inthe country, providing anaverage of over 200,000 tons of fish per year (personal communication, Koranteng, 1994). Pelagic fishery dominated by sardinella, is the most widespread type of fishery in the artisanal sector.* Since the bumper catch of 1972 however,annual yields from sardinella during the major “herring” season (July to October) have been unpredictable causing great concern among al1 interested parties (Koranteng, 1989). Technologicalinnovations that have been introduced at different stages since the late nineteenth century have brought about the remarkable increases in artisanal fisheries yields, and to some degree, modernized artisanal fisheries compared to30 years ago. But certain aspects of the organization andoperations have persisted. Canoe fisheries remain a highly labour intensive activity, and every fishing unit provides on the average two incomes, first to the fisherman, and second to the fish processor. Among the Gas where separate residence for men and women is common this may be viewed asseparate incomes for separate households. Of particular importancein the legacyof the old order in present day artisanal marine fisheries is the continued .dominance of kinship as a basis for labour organization. Previously thought to be a thing that would disappear with 2. Sardinella is popularly known in Ghana as “herrings”. The two main types found here include the round sardinella and the flat sardinella (Koranteng, 1989).
135
modernization, the importance of kinship orgmization in persisting Africm modes of production is now widely recognized. Scientists who are revisiting the Gulf of quinea fisheries resource base now recognise the need ts adopt a holistic approach. They acknowledge that several aspects of the non-biological and technological features of artisanal fisheries need to be more closely examined to improve our understanding and policyfor the sector. Among the issues that cal1 for greater attention is the centrality of the kin group in this business and the social noms, expectations and pressuresthat arise from thatcontext. The objectives of this paperare first to outline ways in which major decisions and pactises in artisanal fisheries are shaped largely by kinship relations, and their implications for change within the sector. Second, the paperaddresses the overlap between power and authority structures in fisheries, andthe wider sociopolitical organization of the fishing communities. For the discussion of these objectives various aspects of fisheries organization in which kinship considerations appear to be quite influential have been identifiedas follows :
- labour recruitment andtraining - ownershiphnheritancehvestment - migrating fishermen: “aprodo”. - authority structure, management and supervision
- cooperative elementsin artisanal fisheries - labour conflicts
- gender relationsin fisheries labour force of kinshiprelations for transformation
- consequences
Information for this paper forms part of a wider study on artisanalfisheries arnong the Ga-Dangmes in the Greater Accra Region, which was conducted at the Institute of Statistical Social and Economic Research (ISSER), University of Ghana (1 988 to 1 990). A lot of the discussionis limited to this regionalcontext in viewsfthe important differences arising from the varying cultural backgrou to be founddong the Ghana Coast. Veryuseful studies that have been undertaken on the socio-economic organizationof Ghanaian fisheriesinclude the works of Lawson (1953), Lawson and Kwei (1974), Vercruijsse (1 984) on the Fantis, Nukunya’s study on the Ewes (1991) andBdotei’s (1991) work on the Fantis.
ackgrounnd of ChDangrne Artisanal Marime Fisheries
The Greater Accra Region of Ghana, traditisnal home of the Ga-Dangmes is quite important in the local production of fresh fish, and until recently was the leading coastal region for the highest number of canses, outboard motors and fishermen. About 25 percent of thefishing communities that have been identified by the Canoe Frame Survey in 1992 are located in the Greater Accra Region (Koranteng et. al., 1993). Many of them are small villages with populationsof less than 5000, but there are also big townson the Coast with populations of over 20,000. From the48 fishing communities that have been recorded between Lanm in the West and Ada in the east of the Region, Ga-Dangme fishermen and fish
processors operate from about 69 landing beaches including Lanma, Teshie, Nungua, Kpone, Old Ningo and Prampram, where the ISSER study collected primary data. Presently about3 percent 1 of marine canoe fishermen the in country are also to be found here (ibid). In production as well the Greater Accra Region is quite important in coastal canoe fisheries. Over 40 percent of the domestic marine catch in 1992 was landed along thisCoast (Fisheries Department, 1994). Poli and ali canoes for pelagic fisheries make up 63 percent of the canoe fleet in the Greater Accra Region, and aatnational level, the region carries over one-third of al1 the poli/ali canoes along theGhana Coast. In additionit has over 50 percent of the line canoes in the country, whichis not surprising giving that Ga-Dangme fishermen are traditionally known for hook and line fishing (Koranteng et.al., 1993). Artisanal marine fisheries is founded on a highly age and gender based segregated division of labour. In Accra this has its roots in a family residential pattern commonly found in traditional Ga families, in which men and women live in multi-generational separate sex compounds. From these compounds and closely related ones a canoe owner or fishsmoker normally recruits young male or female relatives to form work groups. Most of the people involved in fisheries work on the Greater Accra Coast have had little or no forma1 education (over 70 percent). Many of those Who have been to school completed less than sixofyears education. The most educated operators are to be found in the big towns like Teshie, where a few fishermen have completed secondary school. The early age at which boys and girls are recruited into apprenticeships in the industry interferes with their enrolment and retention in school. As in most traditional industries, young people are left with little choice in career selection. Men and women become involved in fishing and related activities as a matter of cause. It is part of the tradition. They remain itinbecause of the general lack of mobility from this sector elsewhere, imposedby their lack of skills (Lawson and Kwei, 1974). Besides, alternativejobs are few in and around many fishing communities. Perhaps this rather than a social stigma attached to fishing as Lawson (ibid) suggests, more realistically explains why fishermen and fishsmokers remainin the sector. In any case crew rnembers wouldit difficult find to combine fishing with other occupations because of their long or frequent periods at sea. The women on the other hand are notedfor being very versatile in their occupational pursuits, often combining fishsmoking with other micro enterprises such as kenkey making or petty trading. Inthe lean season they may concentrate fully on these other trades until fish is in plentiful supply. At Lanma the women are seasonal farmers growing cassava and vegetables during the lean season. Independent fishermen and fishsmokers tend be to older rather than younger members of the compound. It is this group that owns the rneans of production and provides the needed working capital ina work group. This is notsurprising, as one more often than not inherits capital inputs in this occupation from an 137
older relative. Onthe average people stay active in fishing and fishsmolcing wel past their sixties barring ill-health. As they grow older theymove from heavier to lighter tasksS3
cruitment and Training of Work Groing Menabers As can be expected in family based business,a high degree of homogeneity in ethnic composition is found in the work groups (95 percent Ga-Dangmes), even in the large town of Teshie. Apart from the fact that jack of exposure to the sea restricts‘people in the hinterland from entering marine fishing, fishermen generally remainamong their own people. A majority of fishermen believe that one does not needto come from a fishing family to be a fisherman. But their own recruitment experiences suggest otherwise, and they admit that the process of becoming a fisherman or fishsmoker is not very open. About 80 percent or more of the fishermen at Lanma, Teshie and Prampram described themselves as close kinsmen of the other members of the crew on their eanoes. Invariably most of them were related to the owner directly or by marriage. Young boys canjoin canoes on either their mother or their father side; in addition it is quite common for men tofoster their nephews and recmit them for their canoes. Normally the recruitment age is quite low, especially for sons and foster sons who may one day inherit their fathers’ gear. Very Young boys of five to seven yearsolld may be taken on board as helpers, to shove water out of the canoe. These are regarded asthe first steps in training. Fishermen are however not averse to recruiting outsiderstheinto work groups and this happens more oftennow thanin the past . The adoptionof the larger ali, and later, poli and watsa nets have made this necessary.While about fifty years ago the average came crew sizewas about 3 men, it had quadrupled toabout 13 in 1986 and by the time of the study some poli canoes were carrying over 28 crew members. Most crew members learn their skills through apprenticeships with canoe owners. The periods of training tend to Vary between one year in small communities like Lanma, while at Teshie may it extend tofive years, thoughthe reason for this long period is not clear. My impression is that the extended period may be due to the long lineof senior fishermen ahead of the trainees in an intense fishing community like Teshie. It couid also be due tothe age at which as the younger you start the longer the period one is recruited into apprenticeship, of training. But on the average trainees are considered to be ready for full membership of the crew after two years. Typically, apprenticeships are informa1 and therefore “feefree”. In contrast there are relatively expensive formal apprenticeships, in which a father or uncle formally hands over his ward to be trained on a particular canoe. The owner then 3. Retired fishermen or net menders for example, may take up roles as “negotiators” at the beach, where they help the fishermen to bargain for good prices. This is quite common at Teshie.
138
’
specifies the costs that come with this assignment, and the thanks offering that is expected after graduation. There are no fixed amounts that have to be “paid”, but the items usually include alcoholic beverages, and cash. Besides apprenticeshipsthere are no regular and easilyavailable alternative training facilitiesfor preparing fishennen and fishsmokers. Occasionally fisheries extension services organize short orientation courses to teach the use of new technology and techniques of maintenance of outboard motors. Very few fishermen however have the opportunity to attend such programmes.No more than 20 percent of the men we contacted had been at such workshops, mostly from Teshie. Occasionally a full fledged fisherman “applies”to join a Company from another canoe. He may be interviewed by the canoe owner regarding his skills and experience at sea; he is also briefed about the financial state of the Company (debts owed on capital inputs, etc.), andis charged a bottle of schnapps if he is accepted. In the apprentice system of training older men are respected for their experience on the sea, and valued as a very critical resource for the practical training they can provide. However, as the crew gain experience the old are strongly discouraged fromtaking part in fishing trips, although one or two older people on the trip is seen asuseful for maintaining peace, andfor advice during emergencies at sea. But conflicts often arise between different generations in this highly multi-generational labourforce. Because of their advanced age and reduced physical strength, some of the younger crew feel that older men are a liability. In addition theyare accused of constraining othersby autocratic ways of behaviour. Women like their male counterparts tend to rely on their daughters, sisters and other female relatives to form a fish processing work group. Young girls of about five years start running errands during smoking sessions and in this manner they learn fishsmoking and handling fromtheirfemalerelatives.Their responsibilities and experience grow with time and age, until eventually they can take charge of a batch of smoking unsupervised. Unlike infishing training there is hardly ever a forma1 apprenticeship in fishsmoking. None of the three communities reported any such experience. Until the early eighties when IL0 assisted the National Council on Women and Development (NCWD) to organize short training workshopson improved fish processing there were hardly any extension programmes in this area of activity. The key aspect of this project was the introduction of the rectangular multi-tray Chorkor smoker, named afterthe township of Chorkor where it was first developed and tested by the FA0 and the Ghana Food Research Institute (FRI). As this programme was very limited in coverage the majority of fishsmokers we met had never heard of extension services. Prampram however was one of the project sites, and so had benefitted greatly fromthe training and credit facilities that came with the project. Here many women proceeded to build their own Chorkor smokers with assistance from the project.
139
wnership, Imheritanse and Investment The idea of African modes of production such as artisanal fisheries evoke images of subsistence-based primitive communalism, driven by egalitarian distributional forces rather than persona1 interests and the quest for private property. By most accounts any such communalistic systemof production and distribution was already in decline in African political economy well before colonial conquest (Coquery-Vidrovitch,1978). On the contrary there is evidence that indigenous foms of production havefor a long time been characterised by strong market orientation, even where they have retained tkeir kinship-based organization. By bringing to the study of artisanal fisheries an analytical framework that recogniaes its pro-capitalist characteristics one is able to identify various capitalist interests in the investment and distribution of resources in this sector. It is important to take thisstep in order to shed someof the stereotypes that have hindered work in this area.
Widespread misconceptionsin the past that work groups in fisheries operated as de facto cooperatives led to inappropriateassistance programmes that failed to facilitate improvements in the fishing industry. Issues of ownership and ... investment in artisanal fisheries have for a long time been misunderstood by outsiders (Lawson and Kwei, 1974). Among fishermen and fishsmokers at present there are clearly identifiable owners; they are the ones Who have madeal1 or most of the capital investments, and are able to mobilize other key resources to sustain their operations. Vercruijsse (1984) maintains that this represents a significant shift from the past in which the industry apparently was dominated by independent operators involved in hook and line, set net and cast netfishing. Each fisherman moreor less owned his own gear. Labour and capital were thus united in a petty commodity modeof production. As the larger nets and canoes overtook the sma , vessels and gear, fishemen became polarisedinto owners of capital and workers on canoes. A change in the authority structure where less financially independent fishermen had been forced into subordinate positions vis-a-vis their better off '. compatriots, was therefore already noticeableby 1903 (ibid).
A separation of labour and capital appeared with the introduction of the larger ali net and the Mfantse canoe, which required a larger workforce and higher financial outlay (ibid).The changes were not taken lightly. He reportsfurther on protests and clashes between different factions of fishermeneither supporting or denouncing the new technologies.There were strongfears on the opposing side that the new gear would deplete the fisheries resource, and at the same time depress fish prices through increased supply on the market. Many Ga towns including Teshie were involved in the clashes. The conflicts could also be seen as the result of a perceived loss of autonomy and income among fishermen Who could not afford to switchthe to more expensive and higher capacitygear (ibid). The new imbalance in the relationship between awners and workers notwithstanding, it has not been easy tointerpret their relations in conventional 140
’
class analysis as Vercruijsse (ibid) has tried to do. Several complications stand in the way of such an exercise. One finds thatthe Key parties, namely owners of capital and workers, are traditionally found in the same family and bound by several social and economic ties. To a large extent the obligations enshrined in these family relationships competes with other persona1 interests that may arise out of their opposing relationships to the means of production. Given the bonds of kinship and other norms and values shared by the crew and canoe owners, the idea that the crew members and here we might add “assistant fishsmokers” might be considered as
(Assistant Wolei Atse, Teshie, 1989). For this reason young men are encouraged to undertake migrations whenever possible. 5. Migrating Fishermen
Ghanaian fishermen are well known for their migrations to other fishingports in the West African sub-region. Though the Fantis and Ewes have been more active in this area than the Gas-Dangmes, there arecertain fishing communities dong this Coast who regularly undertake such trips. They include fishermen from Kpone, 81d Ningo, to some extentPrampram. In many of the other towns some sporadic migration takes place. Migrations may be for very short periods of a month or less, or very long periods of one year or more. The local system of classification of such movements distinguish between “aprodo” and “hefoo”, where the former refers to long migratory trips and the latter impliesa short spin abroad. Local migrations such as “hefoo” may no longer be as attractive as “aprodo’’ in foreign waters. It is also worth noting that the Ga-Dangme fishing communities no longer attract migrant fishermen as they did in the past, when Fanti and Ewe fishermen regularly came to Chorkor, Jamestown and other places for the lucrative markets. Kinship ties continue to influence the composition of crew members on “aprodo”. Fishermen indicate that it is particularly important when going on “aprodo” to travel with kinsmen or well trusted close friends, because of the potential dangers the crew might face. Young men migrate becausethey find it difficult to save at home. Here there are numerous obligations that oneis unable to avoid. In contrast when on “aprodo” the fishermen save their money to be drawn in bullc at the endof the year. About 90 percent of the fishermen at Prampram and Kpone had been on migratory expeditions in thepast. Their exploitstook them to fishing communitiesin Lome (Togo), Lagos (Nigeria), Abidjan (Cote d’Ivoire) and Libreville (Gabon). Because of the potential dangers and hardships that they can encounter, proud returnees suggest that embarking on “aprodo” is a testof manhood for theyoung fisherman. There is no indication however that, this a formal requirement of initiation into the occupation as a fully qualified member of the Company.
142
During theirstay abroad, the crew continue the same communal style pattern of living that they practise athome. Their housekeeping and meals for example, are usually jointly catered for and the costs are shared by the whole group. Migrant fishermen tend to rely heavily on their kinsmen Who have already settled at the foreign fishing ports.On trips to Abidjan for example, fishermen from old Ningo contact the Ga Wotse (Ga seniorfisherman) in this settlement. He in turn may identify a host from within his own migrant quarter of the community or from the indigenous population among whom they have settled. If the fisherman already have a direct contactthey may contact this person, quite often a woman fish dealer atthe foreign port to serve as their host; this usually involves making arrangements for fishingrights, board and lodging, marketing outlets and credit facilities if necessary (petrol money). Prior information about the situation of kinsmen and women or alternatively, local hosts at the foreign port is therefore critical in the selection of an “aprodo” destination. Women are not excluded from migrations, though they typically join the crew by road rather than accompany them by sea. At Prampram and Kpone about 25 percent of migrant labour in the past had been made up of women. Senior wives of the most senior crew members are allowed a turn on the tripsbefore the other wives. Where egalitarian norms are really strong, the wives rotate so that al1 the wives (or sisters of unmarried fishermen) have a chance tostay with the crew in the foreign country to earn some extra money. They go to provide housekeeping services for themen, for which they receive a shareof the catch or direct wages. Whether the women are able to deal in fish or not at the migrant site depends on the graciousness of the local fish dealers. Not al1 fishermen are in favour of taking women on “aprodo”. They describe the exercise asexpensive. Changes are beginning to take place in the terms under which Ga-Dangme fishermen migrate to other countries. Fishermen by the 1980s were no longer restricted to migrating through the traditional companies of mostly kinsmen. By that time some were leaving Ghana under independent contract arrangements with foreign canoes. A number of fishermen from Old Ningo reported being recruited from here by Gabonese and Beninois canoe owners to work for them in the major’season.Other types of contracts have also developed, which involve the conscious transfer of skills from Ghanaian fishermen to their foreign neighbours. Foreign canoe owners Who come to Accra to buy Ghana canoes, sometimes recruit Ga fishermen totrain their own crew in the use of the canoes in their home countries (Sheves, 1991). Whether such contracts are based on the share system or on fixed wage payments, the fishermen described them as normally very profitable and less complicated than travelling in a Company from home. Whatever the terms under which fishermen work ina Company, outwardly they display a high degree of cooperation which is necessary to make their highly labour intensive enterprise work.
143
6. Gooperative EPements in Canoe Fisheries
Fishermen work on the canoes not as employers and employees as can be found on industrial fishing vessels, but ratheras “partners” with prior agreement on thkir share of the catch. The share systemreplaces fixed wages for the crew. Share arrangements tend to differ for different kinds of fishing operations, and sometimes from one village tothe next. Superficiallythe share system contains elements of an egalitarian cosperative set up. Nevertheless in this cooperativeall men are not equal. On the one hand it casts the crew in the Company as equals of the canoe owner, while at the same time it recognises and rewapds hislher greater investment in the enterprise. The arrangement is probably a residual from thetime when al1 crew members owned their own gear but not al1 owned a canoe; they would therefsre surrender a part of their catch to the canoe owner aftera fishing trip as compensation for the use of the canoe. In al1 only about 3 percent of canoes in the study were owned either by the whole Company or jointly with a canoe owner. Inal1 the communities we visited about 96 percent of poli nets and outboard motors were owned by th canoe owner. On poli/watsa canoes typically used in pelagic fisheries the owner and the .crew share equally the overheads of an expedition, which can be quite substmtial depending on the size of the canoe and number of crew. Typically sverhead expenses are first deducted from the catch. Fuel bills are the highest costs, but the crew also have to finance food, rope and ice as well as sundries such as kerosene. The remaining portionis divided into equal halves; one half is shared by the canoe, the net and the outboard motor which invariably is owned by the same person, whilst the other half is shared amongthe crew. If he accompanies the crew the owner in the share system counted as a member of the crew, andis therefore, entitled to a crew member’s share, in addition to shares awarded to the capital inputs. In this manner the canoe owner hspes to be able to recover _ the costs of hislher investment, and also make enough profit to finance major repairs and spare parts. Where very close bonds exist between theowner and crew the lattermay go to great lengths to help to ease the owner’sdebt burdens. At Lanma a Company agreed to make special deductions from the catch to assist the owner pay off a lsan he had taltento purchase a new net.I-Iere it is common for came owners to forfeit the shares that are normally paid out to the canoe and gear, during the minor season when the catch is low andthe shares of the fishermen are reduced. In contrast some canse owners at Teshie complained bitterlythat the crew had refusedto make any extra deductions to help them pay off their loans quic The crew members felt that they were not receiving any extra assistance from such owners in their times of need, so there wasno need tomalce anyallowances for them. Compassion for the crew when they werefaced with hardship earnsa canoe owner considerable respect and allegiance from their crew. In spite of this high degree of cooperation that is practised by fishing companies, past and present forma1 attempts to form fishing cooperatives al1 over the Greater Accra Region have been disappointing. They have been used 144
mainly.for channelling fishing inputs to the fishing community. As fishing inputs are normally not jointly owned by the crew members they find cooperatives irrelevant to their work. AtTeshie only canoe owners or their bosuns belonged to the cooperative when it was active. A lengthy discussionon the experiments with fisheries producer cooperatives is given by Lawson and Kwei (1974). At present there is a National Association of Canoe Fishermen to which al1 chief fishermen belong. They represent the interests of the fishermen in their communities whether or not they have a registered association. The Chief Fisherman of Tema represents the Greater Accra Region at the national level. Given the weakness of the cooperative effort, leadership in artisanal fisheries has remained with the traditional structures.
7. Authority, Management and Supervision In every community wherefishing is a major occupation it is supported by a hierarchy of offices that form an integral partof the local traditional power and authority structure. The hierarchy of authority in fishing comprises the chief fisherman and his assistants. It is this team that monitors the observance of al1 the local fishing rules and regulations. Though the office holders normally must be fishermen, their positions are not by appointment but attained through inheritance. The chief fishermandraws his council from the canoe owners inthe community. The chief fisherman and his officers govern and monitor affairs among fishermen and fish dealers in their respective communities by local ritualistic norms and values. These typically have practical as well as spiritual significance, and are initiated and regulated by the chief fisherman and his office holders. Some of the most important of these include the ban on fishing on Tuesdays, purification rites that are performed periodically, the blessing of new canoes and the control of migrant fishermen. Migrant fishermen have certain conditions to fulfil before they are allowed to stay in a host village, which often involves the payment of fees to the chief fisherman.The Tuesday ban though considered to be a day of rest, provides a time for meetings among fishermen to resolve problems, disputes and to perform minor repairson their gear. The women also find it a convenient time to transport processed fish to the market. In many ways the chief fisherman and his heirs and council enjoyan elevated status in their communities, in which the ordinary crew membersdo not share. At the operational level most of the management and supervisory duties on a canoe for example are performed by people selected by the owner of the canoe/ gear. In practice these offices are hereditary rather than appointive, though in the absence of a reliable siblingor younger relative, owners can invite outsiders to represent them in management / supervisory capacity. However due to the introduction of modern technology new offices have emergedin the Company which are largely based on meritski11 andrather than kin affiliation. These include the position of motor man and accounts clerHsecretary, quite often they are young men with relatively higher education thanthe rest of the crew. When it comes 145
to technical decisions their expertise is respected by the crew, which in some cases is a constant source of irritation to the traditional leaders inthe Company. A typieal fishing Company crew has the following offices: - Eelenaa Tsee (canoeowner) - Bosun (representative of the canoe owner) - Assistant Bosun - Motor man/engineer (in charge of the outboard motor) - Paado mole (paddle man) - Wolei (crew members) - Bookman (accountsclerk) - Apprentices
Outside direct fishing operations the crew are bound to thecanoe owner in a kind of patronage syste,m, where the owner is expected to provide material assistance from his or her private resources for fishermen when they have emergencies, or on festive occasions. This creates a network of indebtedness between the two parties that force them to overlook some of their misgivings about the fishing operation. Fishsmoking compoundshave far less intricate office designations than can be found in fishing, though here too the hierarchy of seniority is quite strictly observed. Within the compound there are clear norms guiding ownership, smoking operations and sharing arrangements between the smoker and her assistants. In some communities there is aLoonye or “fishmother” but her position is not hereditary; she is selected by her peers for size the of her operations, hard work and perhaps credit facilities she is able to provide for the others. She represents them in an informa1 capacity, and helps to settle disputes among the women, or to organize boycottsagainst the fishermenwhen the needarises. Not many communities today have a loonye, and there is a general feeling that her influence has somewhat declined.
8. Labour Csnflicts In spite of the familial ties that bind crew members and further links them to the canoe owner, antagonism, competition and conflict are not uncommon amsng the erew. Theyare quite often inter-generational confrontationstheinsame team, but thereis a gender dimensionto this as well. Fishsmokers often find themselves in direct confrontation with the fishermen over prices and other terms of trade. Canoe owners are not sparedthese tensions, and increasingly they have been the targets of confrontations withthe crew over their operations. Anything from the system of sharing to the selection of fishing spots and landingbeaches can bring about conflict. At sea these can be quite dangerous and ailefforts are made to contain conflicts for settlement on ground. Another type of conflict which c m have very grave consequences is interCompany rivalry between different canoes; normally this remains at a level of jesting especially on festive occasions, but it can easily degenerate depending 146
on the occasion and the issue at stake.A few liveshave been lost in many fishing towns in Accra as a result of such rivalries.The chief fisherman and his council of elders try to handle these matters on their own. Incriminal cases however, the police are called in. The main line of cleavage though is between the older fishermen on the one hand and their subordinate crew members. Among the women it is the lead fishsmoker (the one Who owns the ovens) against her assistants. Nowand again younger sisters complain of domination by their older sisters or mothers, but smoking compounds appearto be less prone to conflicts than fishing companies. Canoe owners talk openly abouttheir misgivings aboutsome of the trends in activities of companies. Growing suspicion on bothissides gradually undennining the traditional camaraderie that has existed between the two parties. At Teshie there were numerous complaints that the younger crew members had become disrespectful and dishonest. The canoe owners believed that they were being forced to retire early from seafaring so that the crew could interfere with the catch. For the fishermen, keeping the owners out of fishing trips reduces the number of men receiving a share in the catch, which boosts the size of their own shares. The companies have been accusedof far worse crimes in recent times. Several cases of disappearing catches had apparently been reported where the canoe owners suspected the crew of landing their catch at other beaches. At Teshie a fishing Company hadactually been caught landing at Nungua instead of Teshie. Muchof the misunderstanding is centred aroundthe relatively high portion of the catch that eventually goes to the canoe owner (the canoe, the net, the outboard motor), especially in these times of poor catches. If it continues this could lead to a confrontation that may bring important changes into sharing arrangements in the future. Whatever differencesare threatening to break their front, one thing fishermen and their canoe owners in common have is their uneasy with fish dealers Who by custom are almost entirely women.
9. Gender Relations in Fisheries Labour Force The artisanal fisheries sector is heavily dependent on close cooperation between the men and women involved this in industry. Apart from the traditional tasks that are segregated by sex, the two parties havecome to provide strategic support for each other based on their peculiar positions in the production and distribution of fish. To a large extent women in the community rely heavily on their male relatives involved in fishing to supply them with fresh fish for processing. Traditionally, wives and female relatives have the first serve of a fisherman’s share of the fish, before he would sel1 to non-relatives. In the past this carried withit softer credit conditionsfor the relatives than was available to other women, but the economic hardships andthe need for ready cash hasforced fishermen to demand cash payments upfront from al1 their customers including their female relations. This is problematic where the fishermen are indebted to their customers. When they find themselves ina fix such the fishermen sometimes take quite drastic steps to protect their interests. 147
At Eanma the fishsmokers on a number of occasions had to Chase their that the crew had fishermen toBortianor about 2 miles away, when they got hint diverted the catch to this beach where theycould obtain cash paymentsfor their fish from non-relatives.The women were not only infuriated by their breaching custom, but perhaps more for the fact that most ofthe crew members owed msney to the women, which theyhad agreed topay for with their share of fish ! The fishemen on the other hand rely greatly on the women not only to process and markettheir fish but also for much needed creditfacilities. They turn to the gsods, for working capital and for tiding fish dealersfor loans to purchase capital them over the lean season. The situation where women are the financiers has arisen largely because of their prominence in the proeessing and marketing of fish after it has been landed. Invariably these women who provide substantial credit facilities to the companies become their favoured customers, and with time supersede the female relatives in the preferential supply of fish. The “petrol in determining women” asthey are known by the crew come to be very influential the beach price of fresh fish. In a place likeTeshie many of them have acquired their own canoes and have their own companies of fishermen. Webs of indebtedness binding various fishermen to women dealers are thus builtup from which the men find it difficult to extricate themselves, because of their perpetual cash flow difficulties. Many fishermen believe they are cheated by their female trading partners, though they have not yet been able to securealternative and accessible sources of credit to replace these partners. A Company at Kpone that returned from an “aprodo” from Lome without repaying loans they owed totheir petrol woman, were chased bythis dealer to Accra, where she managed to put pressure onthe chief fisherman to help her recover the loans. While the outside world may be convinced that fishermen would have to break their indebtedness to fish dealers in order to make any headway in their standard of living, the fishermen view the situation differently. Many of them believe that inthe present circumstances they cannot do without the support of the women. It isthis fear that undermined the successful implementatisn of the FA8 assisted modern wholesale fish market at Takoradi, establishedin the 1950s. Here fishemen would have been able to auction theirfish to the highest bidder rather thanto their traditional customers at low prices. ~
18. Implications of inship Organization for Artisanal Fisheries Develspment. Dependence on kinship as the organizing principle in artisanal fisheries has several implications for its future development. Though some of these appear to be detrimental to progress there are several strengths ofthe system which have enabled it to survive into these times. Organizing around kinship has been relatively effective for controlling and supervising the fishing Company. The canoe owner has a well established normative base in community traditions that are understood by al1 members of the Company, to guide and support hi& 148
her, as well as arbitrate in cases of conflict. Financially the kinship base has facilitated borrowing and lending among operatorsin the sector, a situation that cannot be taken for granted in the absence of alternative sources of credit in most fishing communities. Many loans given outby relatives are interest free, compared tothe high rates( 50 to 100percent) demandedby professional lenders. Other features of the organization of artisanal marinefisheries however, may pose difficulties in attempts tostreamline the sector to make it more “efficient”. As it continues to operate a closed family-based recruitment policy it is very difficult to introduce modern criteria into this process. A more pertinent issue arises over the ability of canoe owners to keep the size of their companies within reasonable limits when indeed the extended family remainsthe basis of labour supply. Some fishermen atTeshie complained that the companies were too large because al1 the young men in the family expected to work on their father or uncle’s canoe. They blamed this on the lack of employment in general, which gave such young people few alternatives. The young age of recruitment for training trough apprenticeships, thoughit is cheap and accessible, has important drawbacks. It may be partly responsible for the low educational status of most fishermen in Accra and elsewherein the region. Whereas fishermen have adopted technologies that have revolutionalized their movement at sea and the size of catch, their lack of forma1education couldbe the reasonfor the slow appreciation of other basic technologies.The echo-sounder which is regarded as a fairly simple technology for spotting schools of fish has been tried and effectively demonstrated here but fishermen have not showninterest (Kwei, 1973). Haakonsen’s (1988) analysisof the prospects for young peoplein fishing is more optimistic than most. He hasa point about the vibrant nature of the sector, and its impressive contribution to domestic production. But perhaps he relies too heavily on the continuous Stream of labour into the industry as an indication of its prosperity. Whether this expansion can actually be attributed to potential gains to be made in thissector, or it is merely a reflection of limited employment avenues would need further study. Canoe owners have hinted that sometimes they take on board more young men than is necessary because they are in the family. Haakonsen further makes an important observation thatthere is never a shortage of buyers for outboard motors and nets. It would appear however athat significant proportion of buyers tend to be existing canoe owners, awith limited number of new newcomers. Several people have noted the wide gulf between the tremendous increasein the labour force in contrast tothe limited expansion in the number of canoes since 1953.5
5. In 1953 there were an estimated 8000 canoes with 56,000 fishermen. By 1986 there were an
estimated 8,288 canoes with 108,000 fishermen (Odoi-Akersie, 1988).
149
Il. Summary
Artisanal marine fisheries as it operates in the Greater Accra Region and indeed in most parts of Ghana, is still largely organized and sustained by the family. Leadership in the sector is in several instances hereditary rather than appointive, and forms an integral partof traditional authorityin the community. This brings to the work groups strong social obligationsin the way people relate to each other.The effects of this are most pronouncedin the natureof recmitment, authority structure and management, as well as gender relations in theindustry. Fishermen and fishsmokers thrive on mutual assistance in material t e m s and in the form of mutual support. This is practised for the benefit of people within the same work groups, but it also operates between work groups. Coupled capitalist with the strong kinship base however is a significant individualistic and economic orientation which guides the way business is conducted in artisanal in the same work groups fisheries. Based on this kinsmen and women involved are clearly divided into the owners and the workers. Whatever cooperative elements exist in the industry therefore, (and there are several) excludes joint or communal ownership of the means of production. Poor understanding of this fundamental fact has ledfisheries planners in the past ts introduce socialiststyle producer cooperatives with negative results. For almost al1 the parties involvedfishing and relatedactivities is notjust an occupation but a way of life. Although some rnembers of fishing communities have ventured outside to seek employment in other sectors, fishing remains .a popular avenue to young people. Their exly entry into fisheries and long periods at sea precludes their preparation for any other form of occupation, and limits their interaction with the wider Society. Giventhe rather high expenseinvolved in owning poli fishing gear, and the relatively few Who ever achieve this status the average fisheman has little hope of being an independent came owner in his lifetime. Fish processinggear is considerably cheaper to acquire thanfishing gear, which makes it possible for more women to set up on their own. Women are also aided by their versatility in occupational skills to combine fish process with other activities to boost their incomes. A handful of canoe owners and big fish traders can erroneously convey very examined positive images of prosperity in the sector to the onlooker. But closely the standard of living for both men and women in the industry appears to be relatively low. Detailed studies on this aspect are not readily available; it is therefore difficult to state conclusively that fishing communities are poorer than others in comparable indigenous occupations. My impressions of fishing communities however, areof widespread poverty among both men and women These are illustrated in poor housing, limited infrastructure and high levels of illiteracy. The situation appears be to the same inboth the large and smallfishing communities.
REFERENCES Coquery-Vidrovitch, C. (1977) Research on the African Mode of Production, in P.C. W. Gutkind and P. Waterman (eds.) African Social Studies, London, Heinemann. Fisheries Division, (1966-67) Annual Report, FisheriesDivision, Ministry of Agriculture, Ghana. Fisheries Department, (1994) Summary of Marine Fish Production in Ghana, Research and Utilization Branch, Tema, Ghana. Haakonsen, J.M. and M. Chimere Diaw (1991) Fishermen’s Migrations in West Africa, IDAF/WP/36, FAO/DANIDA/ Norway.(Contains articles by Nukunya and Odotei) Haakonsen, J.M. (ed.) Recent Developments of the Artisanal Fisheries in Ghana, IDAFWPI21. Cotonu. Koranteng, K.A. (1 990) Ghana Canoe Frame Survey, 1989, Information Report, No. 25, Fisheries Department, Research and Utilization Branch, Tema, Ghana. Koranteng, K.A. (1989) The Sardinella (Herring) Fishery in Ghana: The Past,Recent Developments and Years Ahead, Information Report, No. 23, Fisheries Department, Research and Utilization Branch, Tema, Ghana. Koranteng, KA.; 0.0.Nmashie; A.N.A. Baddoo (1993) Ghana Canoe Frame Survey, Information ReDort, No. 28, Fisheries Department, Research and Utilization Branch, Tema, Ghana. Kwei, E.A. (1973) Annual ‘ReDort on the Fisher! Research Unit, Fishery Research Unit, Tema. Lawson, R.M. and E. A. Kwei (1974) African Entrepreneurship and Economic Growth: A Case Studv of the Fishing Industry in Ghana, Accra, Ghana Universities Press. Odoi-Akersie, W. (1988) A Preliminary Analysis of the 1986 Canoe Frame Survey, in J.M. Haakonsen (ed.) Recent Developments of the Artisanal Fisheries in Ghana, IDAF/WP 21. Cotonu Vercruijsse, E. (1984) The Penetrationof CaDitalism: A West African Case Study, London, Zed Books.
151
Abstsaet :
The population genetics of marine fishes has long sufferedfrom the absence of spectacular results which leda lack to of confidence fromthe fishery managers. The genetic differentiation of two sardine species (Sardinella aurita and S. maderensis) has been looked atwith allozymic data. ForS. altriter we could not ,. find enough genetic variability to infer genetic structuration. S. maderensis showed morevariability. We could thereforetest the possible differentiation of this species in its distribution area. We found a small but significantly different from zero (i.e. from panmixia) value for theWEIR and COCMEWHAM (1984) estimator of WRIGHT’S Fst index of genetic differentiation (Fst = 0.0085). Genetic differentiatisn exists for Somatlerensis.This genetic differentiation means that it is also a dynamic one. We can estimatethe number of effective migrants between the different populations. This number (around 30) is probably negli compared tothe effective size of the populations.We neednevertheless estimates of the effective size to better understand the genetic structure of pelagic fishes. Deficits of heterozygotes inside the samples (Fis values between 0.10 and 0.19) have been observed. This could be due to the existence of different inside a school, relatedness of individuals sampled, or selection effects. We discuss these points briefly. r-
152
Résumé : La génétique des organismes marins a longtenzps souffert du manque de conflance des gestionnaires des pêches par l’absence de résultats spectaculaires. Nous avons analysé la variabilité génétique de deux espèces de sardinelles (Sardinella aurita et S. maderensis) par l’internzédiaire de sa variabilité allozymique. Nous n’avons pas trouvé de variabilité sufisante chez S aurita pour pouvoir conclure sur sa structuration génétique. En revanche S. maderensis s’est révélée plus variable. Nous avons donc pu chercher à mettre en éviderzce sa différentiation génétique sur son aire de répartition. La valeur de l ’indice Fst de WRIGHT estimé selon WEIR et COCKERHAM (1984) s’est révélée faible (Fst = 0.0085) mais significativement ditérente de zéro (c’est à dire de la panmixie). Une différenciation génétique existe donc pour S. maderensis. Cette différenciation génétique est donc également dynamique. Nous avons en effet pu estimer le nombre de migrants efficaces entre les différentes populations. Ce nombre (autour de 30) est vraisemblablementnégligeable devant la taille eficace de lapopulation. Nous avons cependant besoin d’estimation de cette taille eficace pour nzieux comprendre la structure des poissons pélagiques. Des déflcits en hétérozygotes ont été observés à l’intérieur des échantillons (valeurs de Fis comprises entre 0.10 et 0.19). cela peut être dû à l’existence de ccpopulations>> à l’intérieur des bancs, à un apparentement entre individus ou encore à des phénomènes sélectifs. Nous discutons brièvement ces points.
1. Introduction The population genetics of fishes and marine organisms in general has not been as well developed as that of other organisms. This relatively small number of studies in marine fish was probably due to a lack of confidence from fishery managers (UTTER, 1991). The first works on fishes tried either to find simple relationships between productiontraits and one or two genes, or to find markers which wouldhelp discriminate different populations (UTTER,1991). If the first hope has been abandoned the second one is still alive. Nevertheless the latter has lost part of its simplicity: it has indeed been shown that phenotypic differences (size, weight, numberof vertebra, etc.) cannot be considered as good markers of genetic differentiation (HEDGECOCK et al., 1989 ;ALLENDORF et al., 1991; UTTER, 1991). Whatever the genetic markers used to discrimate two or more populations they need to be polymorphic enough to allow differences between these populations to have arisen. If this marker is monomorphic it will be unable to answer any question about populationstructure. If methods of mark-recapture seem to be appropriate to understand how different regions exchange individuals, they suffer a strong bias because they only give information on movenzent of individuals not on their reproductive
153
s the definition of a population is implicitly or explicitlya genetic one (OVENDEN, 1990 ; MARCHAL, 1991b ; B W T and MARCHAL, 1992) the information on the movement of individuals is not sufficient. Only genetical studies which are based ona marker thatis affected by the history of a species is able to account for that kind of event. Of course genetics could not be a useful tool without the different approaches of other disciplines such as biogeography, ecology, dynamics and sclerochronology (for the latter see PANFILI’s project in this volume), etc. The pattern of variation in genetic markers will bethe result of history and therefore of many different forces and events, past and present (see appendix). Our work followed a classical route in that we decided to first work on enzymatic variability. This genetical toolis indeed the easiest one to develop on a new species and permits to work on a great number of fishes on many different loci. Only three studies on Sardinella aurita had been published as we began Our work. BARON ( 1971, 1973) workedon esterases and transferrinesof the blood and was not able to see anough variability. WILSON and AEBERDI (1991) addressed a somewhat different problem since they tried to show that both Sardinella aurita and S. brasiliensis species werethe same (S. aurita) at leastin . the eastern Gulf of Mexico. Only one study seems to haveever been published on S. maderensis (BARON, 1971).
2. Materials and Methsds
The fishes were caught dong the African coast during the year 1992 at four different sampling points: Dakar (Senegal : 62 aurita + 95 nzaderensis), Abidjan (Ivory Coast: 113 aurita + 126 ~ ~ a r l e v e & ~ )Tema , (Ghana: 1 17 aurita + 1 15 maderensis) and Pointe Noire (Congo: (98 + 92) aurita + 69 maderensis). The samples from Ivory coast and Ghana were obtained by pooling individuals of different sampling points off their respective coasts. There were two sampling . dates for S. aurifa off Pointe Noire (98+ 92). The fishes were either immediately dissected or frozen first at -20°C. Eyes, . white muscle and liver were then taken from every individual and stored at 20°C before being thawed and homogenized in order to be used for the electrophoresis. We used strips of filter paper which were dipped in the centrifugated homogenized extract, blotted on filter paper and then applied to the starch gel. The extracts were then refrozen and kept at -20°C for subsequen use. The gels were mn between three and seven at hours intensity settings between 50 and 100 mA depending on the buffer system and other practical consideratio WRIGHT’S fixation indices (Fit, Fst and Fis) are basic parameters in the study of population genetic differentiation. The Fit index measures the departure from panmixia in the total species’ range sampled. If the population is panmictic it is said to be at the HARDY-WEINBERG equilibrium. This equilibrium can be definedby the expected proportions of both homozygotes and of heterozygotes This total departure from panmixia can be partitionned in two terms. The first one is thedeparture atthe population level.It is then measuredby the Fis index 154
either as an excess or as a deficit of heterozygotes. The second term measures the differentiation of the total population intosubpopulations.It is the Fst index. These indices satisfy the identity :
-
-
1 - FIT = (1 FIS) (1 FST)
(WRIGHT, 1951)
As we are concerned with problemsof differentiation between samples we will mostlyfocus on the Fst index.Its value are between zero.and unity. If Fst is equal to zero there is no subdivision the at level observed. If Fst equals unity the different populations are fixed for different alleles. Actually Fstis a standardized variance of the allelic frequencies. It represents the part of the allelic frequency variation which is due to differentiation. Its departure from zero is therefore a measure of the subdivision of the populations. Fis and Fit both Vary between minus one and one. Different estimators of these indices have been proposed (COCKERHAM, 1969 ;NEI, 1975 ;WEIR and COCKERHAM,1984) taking or not intoaccount more than two alleles per locus and one or many loci. Even the comparative values of these estimators have been studied (WEIR and COCKERHAM, 1984 ; CHAKRABORTY and LEIMAR, 1987 ; CHAKRABORTY and DANKERHOPFE, 1991). The data on allele frequencies were analysedusing the GENETIX program developed by the in Montpellier (France). The method of WEIR and COCKERHAM (1984) was used to calculate the Fstatistics for several alleles at a locus to estimate the level of population differentiation at polymorphic loci. We used then the averaged value over the different loci as proposed by WEIR and COCKERHAM (1984) to have a single estimate of the structure of the species on the sampled area. To estimate the departure of the Fst value from zero (i.e. from panmixia),the GENETIX program is generating permutations of the allelesof each locus creating this way a panmictic urn which represents what should happen if the individuals were reproducing randomly and not like more or less independant populations. For each permutation the Fst (and Fis andFit) value is calculated. After a great number of permutations a distribution of the Fst is obtained. The actual valueis then compared to the distribution. The percentage of values of the distribution higher than the actual is obtained. It is then possible to see whether or not the actual value is statistically different from zero. The number of permutations depends on the size of the data file. It was enough to perform 3000 permutations for Our data. Other sets of replicates weredone and never differed from the initial results.
155
Sardinella aurita Among the 15 loci that we could read and interpret correctly only two were variable. We could not therefore use most of them to answer the question of genetic structuration. The reader should beaware that this does not mean at ail the thatspecies can be regarded as a single panmictic unit. It only means that another genetical tool should be used tosolve the problem. WILSON and ALBERDI (1991) also found a very low degree of genetic variability (at the same time they noticed problems to read some of the loci studied). Their study was based on 36 to 41 loci but could not find significant allelic differences between samples of the Gulf of Mexico and the sample of Brazil.
Sardinella maderensis We found five variable and interpretable loci for S. maderensis: ACP, SDH, FDPl ,FDP2 and 6PGDH.
At the African Ievel The Fst value is 0.0085. This value indicates a very lowdegree of subdivision. But the permutations generated by GENETIX showed that although low this value of Fst was significantly different from zero. We present in fig.1 the distribution of the valuescalculated after each permutation. We can see that the value of Our original dataset is in the extremeside of the distribution. This means that there is a population subdivisiondong the African Coast for S. maderensis. We present one distribution for which 1000 permutations were used. We also used simulations increasing the number of permutations up to 5000. We made 1.6 9% replicates for any given number of permutations and never had more than of the values superior to our real value. As Fst is 0.0685 thisdifferentiation seems quite small and we shall discuss it further.
At the Ivory Coast- Ghana b v e l The same analysis gave a value of -0.000’7 for Fst. We must first notice that this negative value is inherent to the parametric mode1 proposed by COCmWARII (1969). See C O C K l ? W M (1969,1973) andCHAKIPABORTY and DAIVKER-HOPFE (1991) for further discussion. When such a value is obtained it meansthat its expectation is zero. The permutation analysis indicated that this value is not significantly different from zero. There is a priori no reason to think that the ivorio-ghanean stock of S. maderensis is composed of two (or more) reproductive units. Only further analyses could help us answer this question. 156
At the population level: the large Fis values As we are much more concerned with genetic subdivision, the following results will be presented rapidly. The Fis measures thedeparture from panmixia inside eachsample. There is one value for each sample although it is possible to have an averaged value over the samples. The WEIR and COCKERHAM (1 984) estimator gives such an averaged value. If its value is significantly different from zero it means that inside the sample the individuals are not the resultof a panmixia. Positive values correspond todeficit of heterozygotes while negative values are caused by excess of heterozygotes. Al1 Fis values were positive and varied between O. 10 and 0.19 (values obtained for al1 population pairs). They were clearly different from zero as no Fis value obtainedby the permutations was larger than the actual. 4. Discussion
Sardinella aurita If we could not yet answer the questionof subdivision of that species along the African Coast, we must pointa few thingsof importance. First of al1these results do not mean thatthere is no subdivision.We just are unable so far to answer this question for this species with this set of genetic markers. We need another genetic approach which will provide more variability. Second the lack of genetic variability that has been observed confirms the results of WILSON and ALBERDI (1 991) as we already noticed earlier. This result raises a problem which cannot be neglectedin the studyof pelagic fishes as it gives information aboutthe possible history and reproductive behaviour of that kind offishes. How can indeed such a species showso little genetic variability if it composed of so many individuals? The loss of genetic variability is al1 the more probablethat a species is composed of few individuals and therefore subject to genetic drift. Rare alleles are then lost randomly becauseof the small number of individuals reproducing. There seems to be a contradiction: on one side you have manyindividuals and on the otherside you havelittle genetic variability at loci which are neutral. Therefore we shall introduce here an idea that should help understandthat phenomenon. In population genetics there is a concept which tries totake account for the real numberof individuals which reproduce. It is the effective size. This effective size (Ne) is of course smaller than the total number of individuals. It takes into account the unequal numbers of males and females, the variance in the reproductive success, overlapping generations,etc. The problem is of course to estimate the effective size of the population. It can be approachedby temporal changes inallele frequencies (WAPLES,1989) but aecessitatesa few generations to provide good estimates of Ne. Whatever the accuracy of this methodto quantify Ne, Our results indicate clearly thatthe effective size of S. auritu is muchsmaller than expected by the catches data. 157
A few hypotheses can be advanced to explain it.This Iow degree of genetic variability can be due to bottleneck effects which can have occured about 18 000 years ago. The sea was about 1O0 meter below its currentlevel. This implies that there was a dramatic reduction of the continentalshelf. As S. aurita seems to be dependant on this environmentit could have suffereda strong reduction of its distribution. If this dependance can be proved (which has not yet been done) there seems to have been two possible refugia,one at the Bissago islands where the continental shelf is very large and the other off Angola. The study of scale deposition by core sampling couldbe of great help to lrnow the past distribution of S. aurifa along the african coast(see the interesting studyof SHACKELTOM (1 986)on scale deposition of pilchard and anchovy off Namibia). Nevertheless anoxic conditions are necessary for the conservation of the scales (SBUTAR and ISAACS, 1974 ; SHACKELTON, 1986). It seems difficult , if possible, to find that conditions in the West African coast. The lack of variability can also be caused by regular varietions in the populations of S. aurita. It is indeed well known that this species is dependant on upwelling systems and sensitive to environmental variations (CURY and FONTANA, 1988;PEZENNEC et al., 1993). It exhibits mnch variability in the catches in very small time scales (a few years). Again data from core samples could be of interest to study thep s t (and long-term) variabilityof this species. Another factor that can be responsible for a small effective size in pelagic fish species is the variance in the reproductive success. A female of S. aurita (and of S. maderensis) is able to hatch between 50 000 and 200 000 eggs per year (CONAND, 1977). Thereby a few number of females can give rise to a population (at least a great numberof individuals) if the environment conditions were good for the larvae produced. This variancecan be even greater because the environment is variable and can induce strong mortality variability.
ardinella maderensis At the Af%xm bved Though this species shows more variability than S. aurita, we cannot generalize this result as we worked ona small numberof loci. Nevertheless this variability was large enough to show that along the African coast the species cannot be considered as a single population. This result is not a minor oneas we know that the triggerfish (Balistes earolinensis) populations have invaded very large areas in a few years (CAVERIVIERE, 1991 ; MARCHAL, 199l a ; PEZENNEC et al., 1993). Indeed the abilityof some fish populations to invade quickly large areas can lead toa genetic uniformity on that area. Our observation of different levels of genetic variabilityof thetwo sardine species is nevertheless c.onsistent with the catches of the two species. The fact that S. nurifa is economically much more important (thecatches are significantly higher) does not mean that its effective size (Ne) is larger. Indeedthe effective size along the
158
time is not the arithmetical mean of the population effective size Say each year but its harmonic mean. The lowest levelof Ne reached during time plays the major role as it is responsible forthe loss of genetic variability. If a species reaches a very high level after a crash the variability will not be increased by this way as new alleles will not be created so fast. S. auritu appears to collapse much more than S. maderensis and frequently has population sizes less than that of S. maderensis. This indicates a (presumably) larger effective size for the latter species. In fact the Fst value, small but significantly different from zero, means that the differentiation is not very important from a genetical pointof view. This can be due to continuous exchange of individuals (migration) between the populations. This can also be the result of a recent separation: the allelic frequencies would not have had time enough to diverge by genetic drift. Al1 the more that the effective size might not be very small. The fishery managers are concerned with populations of fishes over very short periods of time. The studies of population genetics on marine fish have been rarely able to prove the existence of differentiation even on large areas while it has been observedfor freshwater species (GYLLENSTEN,1985). This kind of result has been largely considered as a failure of population geneticists. WYATT et al. (1991) consider indeed that the existence of these units of reproduction is a fact but that the lack of variability and the possible time scale inferred has been too short. Thereby JORSTADet al. (1991) have been calling for other genetic tools. Fishery managers should notforget that the period of time examined by the geneticist is larger thana few years. But this implies that when a geneticist shows that there is population subdivision this result is likely to be valid. Moreover under a few hypotheses theFst value can be relied to estimate an of the number of effective migrants ( M ) between the populations : Fst = 1/(1 + 4M). (CROW and KIMURA, 1970) Our valueof Fst is 0.0085 thereforeM is about30. Our results meantherefore that, as a result of their history since they separated, the current population differentiation'is equivalent to that they would haveif they had exchangedabout 30 individuals every generation. These migrants are effective migrants. The populations could exchange thousands of individuals but only 30 could reproduce successfully. This result is of a great importance as it underlinesthe difference between the dynamic as opposed the to genetic evolution. The value of M is only approximative and cannot of course be used as an absolute value as long as we do not have ideasof the effective size of S. maderensis. Nevertheless this value is certainly negligible as compared toNe. We have thereby twoanswers at two different levels. The populations have genetically diverged. This first point is necessary to stress the second one. As these populations have significantly diverged we can estimate the value of M which shows that from a dynamical point of view the exchange of migrants is small. Otherwise we would not find any significant Fst at all. 159
No subdivision has been proved to exist. But the useof another markercodld be of interest aswe lknowthat even at such a large scale as the African distributi (between Dakar and Pointe Noire) the differentiation is not very marked. The existence of different populations cannot be denied by Our results. The use of another genetic tool is of importance. If the mitochondrial DNA proves to be variable it could be very usefulfor understanding the population subdivision of this fishery and therefore its managing. e' the
la^^^ bvel :the large Pis values
First, we must stress the fact that the individuals sampledare members of a '. school as they have been bought to fishermen on the beach (mostof the time and wh& not they were taken froma case of a boat). We can therefore suppose that they belong either to one school or to the pooling of a few schools. According to SHUSTER (submitted and pers. comm.) if the Fis values are more than as twice large as Fst values type, smaller than the ,is used to set the beach seine. * the season, importers should bring i n their frozen sardinella outside the season - from January to June, a period of high demand but low domestie supplies. As far as international trade is concerned, smolted sardinella prices are also ,, affected by eurrency fluctuations. This is particularly true for the GhandTogo trade. Withthe steady depreciationin the valueof the cedi, the Togolese buyers during the were finding smoked sardinella cheap on the market even offseason whenlocal prices were high.The Ghana/T.ogo/Benin trade is expeeted to significantly expand when situation in LomC is normalized.
422
BILIOGRAPHY FA0 (1992) : Report of ad hoc Working Group on the Coastal PelagicStocks of the Western Gulf of Guinea (Côte d’Ivoire, Ghana, Togo andBenin) CECAFECAF Series 91/56. FAO. Rome Koranteng, K. A. & O. O. Nmashie (1987): A report on the1986 Ghana Canoe Frame Survey. Information Report No 21. Fisheries Research & Utilization Branch, Tema, Ghan Koranteng, K. A., O. O. Nmashie & A. N. A. Baddoo (1993): Ghana Canoe Frame Survey. Information Report No 28. Fisheries Research & Utilization Branch, Tema, Ghana Kwei, E. A. (1966): The fatcycle in the sardine (sardinella aurita) in Ghanaian waters. pp 269-275. In Proceedings of the Symposium on the Oceanography and Fisheries Resources of the Tropical Atlantic. Abidjan, Côte d’Ivoire Olivier, Pezennec (199 1): SmallPelagic Resources and Potentialin the Gulf of Guinea. pp 23-29. In Proceedingsthe of1 st World Conference on Small Pelagics, Abidjan, Côted’Ivoire Tettey, E. O. & K. Klousseh (1992): Transport of cured fish from Mamprobi, Accra (Ghana)to Cotonou (Benin): Tradeformalities and constraints. EEC Regional Programme to Improve Postharvest Utilization of Artisanal Fish Catches in West Africa. Fisheries Technical Report Series No 2/92. Abidjan, Côte d’Ivoire Tettey, E. O. & N. N. Zigani (1992) : Smoked sardinella export trade from Abidjan (Côte d’Ivoire)to Ouagadougou (Burkina Faso): Rail transport provides a vital link. EEC Regional Programme to Improve Post-harvest Utilization of Artisanal Fish Catches in West Africa. Fisheries Technical Report Series No4/92. Abidjan,Côte d’Ivoire
423
E 1:LANDINGS OF SA ITA IN THE SUB BY COUNTRY AND FISHER%",1980-1992 (1.000MT)
1
1980
1
1.83
1981
8.73
1982
112 1
1
0.91
1
2.77
1
1
1985
6.20
14.66
7.33
24.39
0.10
1
n/a
1
24.74
1
18.50
72.83
58.57 n/a 1.1
3.40
1
9.43
1
54.07
1
0.61
1
0.12
1
107.12
Source : CRWORSTOM, Abidjan, Côte d'Ivoire; FBUB, Tema, Ghana; FAO, Rome, Italie
(*) Bénin landings are not included in totals for years 198O/84.
424
1
40.38 n/a
0.98
36.30 9.02 0.47 15.36 n/a
12.04
7.17
1
1
n/a 0.2726.18 10.08 5.53
11.68 1983 1984
19.13
-
1
ANNEX II: LANDINGS OF SARDINELLA MADERENSIS INTHE SUB-REGION, BY COUNTRY AND FISHERY, 1980-92 (1,000 MT)
Year
Côte d’Ivoire
Ghana
Togo
Benin
Ind.
Art.
Ind.
Art.
0.65 0.86 0.87 0.61 1.O3 1.83 2.09 1.70 0.10 0.32 0.94 0.25 O. 17
11.31 12.45 14.01 7.67
1980 1981 1982 1983
8.53 9.88 4.82 8.15
1984 1985 1986 1987 1988 1989 1990 1991 1992
7.97 7.42 10.08 9.18 6.41 7.72 11.60 7.76 2.08
0.35 0.67 0.62 9.06 3.18 7.65 8.70 9.00 2.27 2.12 4.91 2.14 2.39
*Total
0.11 nla20.95 0.06n la23.92 0.09 nla 20.41 0.42 nla 25.91
10.08 1.19 nla23.45 22.23 0.5 1 0.9540.59 16.63 0.481.3139.29 25.48 0.452.4948.00 10.45 080.92 20.26 14.10 nla nla nfa 14.55 nla d a nla 8.21 nla nla nla 15.14 nla nla nla
Source : CRO/ORSTOM, Abidjan, Côte d’Ivoire FRUB, Tema, Ghana FAO, Rome, Italy
* Benin landings are not included in totals for years 1980-84
425
SOMAL MBVEMENTS
D. PAUEY
'
enter for Living Aquatie esourees Mangement (HCLA M.C.P.O. Box 2631, 8718 Makatii,Metrs Manilla, Philippines. Abstract : ~~
~
Three speciesof sardinella (S. aurita, S. maderensis and S. rozai) occur in the waters of Sierra Leone, West Africa. Previous publications suggest that these sardinella migrate part of the year to neighboring Guinea, thus being unavailable to craft fishing off Sierra Leone. Off Northwest Africa, the great change of sea surface (SST) temperature appears to be main the factor regulating fish migration; however, in Sierra Leone SST changes are small, thus there is not the same environmental signal for sardinella to move north- or southwards in order to remain withina given temperaturerange. Analysis of data from Russian research surveys undertaken during the years 1976 - 1990 off Sierra Leone reveals, on the other hand. a strong patternof seasonal onshore - offshore abundamce,wkich coincide with the alternance of the rainy and dry seasons. Decrease in salinitgrin the near-shore waters caused by fluvial mn-offs is likely to bethe main reason for sardinella to move todeeper offshore waters. These findings have consequences for the managementof the sardinella stocks off Sierra Leone, which may not be shared with neighboring countries, thus rnaking their management less dependent of decisisns taken outside of Sierra Leone.
Trois espsptces de Sardinella (S. nurita, S. maderensis, S. rouxi) existent dans les eaux de Sierra Leone (Afrique de l'Ouest). Des travaux antérieurs suggGrent que les sardirzelles migrent une partie de l'année en Guinée voisine., devenant donc indisponibles aux petites embarcations de pêche du Sierra Leone. AU large de la partie nord-ouest de 1 'Afrique, les changements considérables de a) Submitted to Dr. F.X. Bard, and K. Koranteng Scientific Editors of DUSRU Acts
426
température de la meren surface (SST)parait être lefacteur principal gouvernant la migration des poissons. Toutefois en Sierra Leone, les changements de SST sont faibles et donc il semble que les sardirzelles ne reGoiventpas le même signal environnemental qui les fasse migrer vers le nordou le sud pour rester dans les mêmes plages de température. L’analysedes données collectées par les croisières de recherche russes pendant les années 1976-1990 au large du Sierra Leone, montrent au contraire un clair schéma de répartition de l’abondance selon les saisons entre la zone du large et la zone côtière, qui coïncide avec 1 ’alternance des saisons sèche et humide. L’abaissement de la salinité dans les eaux cStières . , provoquée par la décharge des fleuves semble la causeprobable du déplacement des sardinelles vers les eaux plus profondes du large. Ces constations ont des conséquencespour la gestion des’stocksde sardinelles du Sierra Leone, qui ne seraient donc pas partagés avec les pays voisins, et donc rendaient k l l r gestion moins dépendante de décisions prises à l’extérieur de ce pays. 1. Introduction Sardinella form an important part of the Sierra Leonean fishery resources and are represented by three species:Sardinella maderensis, S. aurita and the less important S. rouxi. The artisanal fishery mainlyexploits Sardinella maderensis which is caught with ringnets laid out from planked canoes the in shallow nearshore waters. Though reliable figures on total catch are not available, it is estimated that some 20,000t of sardinellas are caught annually by the artisanal fishery. The industrial fishery operates further offshore and targets S. aurita, though S. maderensis and to a lesser extent S. rouxi are also found in the catch of pelagic trawls and purse seiners. Overall, the purse seiners operating in Sierra Leonean waters in 1991 reported a sardinella catch of 8,500 t (Department of Fisheries, unpublished data). A review of the migration patternof sardinella along the West AfricanCoast by Garcia (1982) suggested that Sardinella nzaderensis performs anti-clockwise migration along the Coast of Sierra Leone and Guinea, which would make them largely unavailable to the fishery off Sierra Leone during the months of June to August. Garcia (1982) was less specific about the migrations of S. aurita and suggested two possible patterns for the stock occurring off Sierra Leone: i) itis part of a larger stock witha distribution rangingfrom Guinea Bissau tothe Sierra LeoneadLiberian border in the south and ii) it represents a separate stock (), awith northern limit in Guinean waters aand migration pattern similar to that of S. maderensis, the existence of which would require joint management between Sierra Leone and Guinea. The aim of the present studyis to test hypotheses derived from these proposed migration patterns, using data on the spatial distribution of Surdinella spp. in Sierra Leone waters generated during 14 yearsof Sierra Leone/USSR research surveys carried out between 1976 and 1990. I
427
2. Material and Methods
Temperature Monthly time-seriesof average sea surface temperature(SST) were extracted from GOADS (Comprehensive Bcean-Atmosphere Data-Set) for the years 1950 to 1990, consisting of weather observations collected by merchant ships over was made withthe help the avorld oceans (Woodruffef al., 1987). The extraction of the database and software developed for the GEBS program (Bakun et al., 1993) for the area dong the West African Coast from 22"N to 7"N (Fig. 1).
bundance Regular resource surveys of the Sierra Eeonean stocks were carried out from 1976 to 1990 in the frameworth ofan agreement between theRepublic of Sierra Leone and the then USSR; the latter provided a research vesse], while the scientific crew was a joint Sierra Leone/USSR team. The surveys were carried out with both bottom and pelagic trawls; the bottom trawl hauls usuallylasted for half an hour, while the pelagic hauls were of varying duration. Data recorded for each haul/station were: vessel's position, time and duration of fishing haul, and - for size range, sex, and maturity each species separately - total number, total weight, stage. The original data sheets were kept at the InstituteO Marine Biology and Oceanography (IMBO) of Fourah Bay College, the University ofSierra Leone, Freetown. A joint IMBO/I@LARM project was launched in 1991 to analyze these data, which were transferred into a relational database and a compter-based Lowlevel Geographic Information System (LE-GIS) ealled SIERRA (Valtily 1992). The available relative abundance data for each species were grouped by season, as follows :
1. January 2. April 3. July 4. October
-
March - June - September - December
Intemediate season, dry Hot season Rainy season Intermediate season,wet
Because the surveys were carried out with different gears, a direct use of catch per hour data as index of abundance would not have been appropriate. : Thus, these data were standardized using the following procedure a) For each cruise, the data were grouped by gear (if more than one gear was used) and by species ; b) In each gearkpecies group the range between0.01 kg and the maximum catch per hour was divided intoclasses six of equal size,ranked from 1 to 6 ; c) In each geadspecies group, the individual catch per hourrecords were in (b) andthe corresponding classified according to the grouping defined class rank assigned to the record ; 428
d) For each cruise, al1 stations where fishing took place but no sardinella catch was made were assigneda rank of zero; e) For each sardinella species, date and geographical reference (latitude and longitude at start of the trawling station) were used to extract al1 records that fell within a defined season(irrespective of the year), and arrange them within a 15 by 15 nautical minutes grid; f) For each season, the values obtained under (c) and (d) were averaged within each grid as defined in (e) to obtain a species-specific mean ranks, used here as indexof relative abundance. 3. Results and Discussion
The mean annual sea surface temperature (SST) offSierra Leone as derived from COADS (Fig. 2) is 27.7"CYthe seasonal amplitude is small, less than 2°C. The annual cycle is characterized by a balanced semi-annual pattern, closely related to the alternating climatic influence from the North during winter and from the South during summer. Maximum values are observed in May and December and minimum values in February and August.The latter correspond to the prevalence of the Sahara-born wind and the height of the rainy season, respectively. Figure 3 represents a time-latitude diagram of the mean annualcycle of SST along the West-African Coast from 7"N up to 23"N, illustrating the prominent variation of SST at higherlatitude compared toSierra Leone (7"N to 9"N); also shown on Fig. 3 is the seasonal distribution of Sardinella aurita along the SenegaleseMauritanian Coast, based on data of Garcia (1982) and Cury & Fontana (1988). It depicts a migratory route that is closely linked to the annual SST cycle, and which results inSardinella aurita being absentfrom certain part of the Senegalese/MauritanianCoast during at least partof the year. If it is accepted that the migratory patterns of S. aurita off Senegal and Mauritania, where seasonal variationof SST are very large, are largely the result of attempts by the fish to remain within a range of preferred temperature (Longhurst and Pauly 1987), one can infer that the low amplitude of seasonal variation in SST off Sierra Leone would exertslittle pressure on the local stock of S. aurita to undertake large migrations. This isexemplified by the straight line in Fig. 3 at 8"N latitude, representing the stationarity of S. aurita off Sierra Leone. This stationarity would not submitthe fish to changes in SST larger than experienced by the migrating fish at higher latitudes. S. maderensis is a coastal species occuring ina broad rangeof temperatures but sensitive to salinity change: it leaves the shore when salinity drops below 3.5 % (Postel 1960, Cury and Fontana 1988).
.
The coastal waters of Sierra Leone are divided into two distinct environments by Sherbro Island, 7"30"N, about mid-way between the Guinea and Liberian borders. The northern part represents a broad shelf, 40 to 70 miles wide; the major river systems of Sierra Leone discharge their watersinto this area. During 429
the rainy season (July to September), about 250 cm of rain fa11 along the coast. The estuaries become diluted and extrude low salinity wedges which are held alongshore by strong southwesterly winds. During the dry season, especially when the dkmnattan>>prevails, northeasterly winds push the low-salinity surface waters offshore, and this is replaced by higher salinity bottom waters, thus generating a small upwelling almg the northern coastof Sierra Leone. South of Sherbro Island, the shelf narrows down to around30 km. It is less influenced by river discharge during the rainy season, as Sherbro Island forms a natural barrier against someof the freshwater outflowof the rivers further north. During the dry season, the strong Guinea current flowing southeast along the coast could be expected to generate a slight upwelling on its left (shore-ward) flank. Thus, it is not surprising that the analysis of the Sierra Leone/USSW survey data suggests two main, persistent areas of distribution forS. nzaderensis (Fig. 4): North of Freetown, close to the Guinea border and south of Sherbro Island. However, with the onset of the rains, the area of greater abundance of S. tnnderensis shifts from the coast towards the edge of the shelf, as illustrated in Fig 4. During January to March S. maderensis is evenlydistributed throughout the shallow waters close to the shore, while in April to June, especially in the northern part of the Sierra Leone shelf, S. nzaderensis abundances are greater offshore. During the heightof the rainy season, July to September, quite a number of grid squares dong the 200/1000 m isobaths showan elevated density, indicative of the movement of S. maderensis to theedge of the shelf.At the end ofthe rainy season (October to December), the distribution reverts back to its normal with S. maderensis close to the shore. Tt should be noted that the 15x15 nautical minutes squares of Sierra are too large to depict well inshorehffshore shifts of abundance in the southern part of Sierra Leone (south of Sherbro Island), where the shelf is only some 30 km wide, as one grid square can cover the whole area from the shallow waters to the 200 m isobath. S. aurita generally prefers the deeper waters further offshore (Ben-Tuvia 1960, Cury and Fontana 1988). This is also apparent in the results of the Sierra Eeone/USSR surveys, which show S. artrien to be more abundant in the deeper waters along the 40 to 200 m isobaths (Fig 5). The distribution differs from S. nzaderensis also in that S. mrit'a seems to be more restricted to the northern part of the Sierra Leonean shelf, being rarely found in large quantities south of Sherbro Island. Similar to S. maderensis, the fairly broad distribution of S. aurita over the whole shelf area north of Sherbro Island during the dry season changes during the rainy season, when S. acwita is mostly found at the edge of the shelf, even though some substantial catches were also made in shallower waters along the Guinea border. After the rainy season (October to December) S. aurita moves closer inshore, especially south of Sherbro Island.
430
5. Conclusions This investigation into the change in abundance of the two major Sardinella species off Sierra Leone found little eyidence thatthese species migrates out of Sierra Leone waters during any part of the year. On the other hand, Our analysis of 14 year’s Worth of survey data reveals a strong pattern of inshore/offshore abundance shifts in close relation to the occurrence of the rainy season.From this we infer locally restricted migrations perpendicular to the Coast, related to the avoidance of low salinity waters and . the utilization of optimum feeding conditions during upwellings. Though it can not be precluded that some of the sardinella in the northern part of the shelf near the northern border of Sierra Leone regularly move into Guinean waters and vice versa, this study does not support a migration pattern implying that the bulk of Sierra Leone’s sardinella would migrate to Guinea (or Liberia) for a part of the year. Notwithstanding that cooperation between neighbouring countries is generally advisable in marine resources management, its practical implementation has often proven difficult. The findings of the present study suggestthat Sierra Leone can pursue the management of its sardinella stocks without having to give much consideration as to what is decided in the two neighboring countries. Acknowledgements \
We thank Dr. Claude Royof ORSTOM for extracting the data in Fig. 1 from COADS, Emile Marcha1 and Philippe Cury for discussions on the biology of sardinella species, and Our colleagues at IMBO for their collaboration. The ICLARMIIMBO project which led to this contribution is funded by the Commission of European Communities, Brussels. This is ICLARM Contribution No. 1058.
REFERENCES Bakun A., V. Christensen, C. Curtis, P. Cury, M.H. Durand, D. Husby, R. Mendelssohn, J. Mendo, R. Parrish, D. Pauly and C. Roy 1993. The Climate and Eastern Ocean Systems project. Naga, 15(4) : 26-30. Ben-Tuvia A. 1960. Synopsis on the systematics and biology of Sardinella maderensis (Lowe). p. 497 - 519 in H. Rosa, Jr. and G. Murphy (Eds.) Proceedings of the World Scientific Meeting the on Biology of Sardines and Related Species, 14 -21 September 1959, Rome, Italy. Vol. 2 Species Synopses. Subject Synopses. Rome,FAO, 623 pp.
43 1
Cury, P. and A. Fontana. 1988. Compétition et stratégies démographiques comparees de deux espbces de sardinelles (Sardinella aurita et Snrdinelln maderensis) des c6tes ouest-africaines. Aquat. Living Resour. 1(3):165-180. Garcia S.1982. Distribution, migration and spawningof the main fishresources in the northern C E C M area. FAO, Rome CE@IAF/ECAFSeries 82/ 25,9 p. Postel E. 1960. Synopsis on the biology of the sardine (Sardinella aurifa Valenciennes) (African Atlantic) p. 55-94 in H. Rosa, JI-.and G. Murphy (Eds.) Proceedings of the World ScientificMeeting on the Biology of M y . Vol. Sardines and Related Species, 14-21 September 1959, Rome, 2 Species Synopses. Subject Synopses. Rome,FAO, 623 pp. Vakily J.M. 1992 Assessing and managing the marine fish resources of Sierra Leone, West Africa. NAGA, The ICLARM Quarterly 15(1): 31-35 Woodruff S.D., R.J. Slutz, W.L. Jenne and P.M. Steurer. 1987.A comprehensive ocean-atmosphere data-set. Bull. Amer. Meteor. Soc., 68, 1239-1250
r~
Figure 1 : Map of North Western Africa, showing theareas (A - H) used for aggregating the sea surface temperatures extracted from the COADS (courtesy of Dr.Claude Roy, ORSTOM). Figure 2 : Mean annual cycle of sea surface temperature offSierra Leone (7"N to 9"N and 11"W to 14"W), derived from the COADS, 1950 to 1990 (courtesy of Dr. Claude Roy, ORSTOM). Figure 3 :Time-latitude diagram of the mean seasonal cycle of sea surface temperature dong the AtlanticCoast of Africa from23"N to 7"N, based on data extracted from the CBADS and averaged by month for the years 1950 to 1990 for the areas in Fig. 1 . The seasonal distribution of Sardinella aurifa off Senegal, describedin Garcia (1 982) and Cury and Fontana (1988), and off Sierra Leone, aresuperimposed. Figure 4 : Distribution of Snrrdinella maderensis in Sierra Leone waters, averaged over three-months periods, from 1976 to 1990, basedon standardized catch-per-effort data from successive Sierra Leone/USSR research surveys. Figure 5 :Distribution of Sardinelln aurita in Sierra Leone waters, averaged over three-months periods, from 1976 to 1990, based on standardized catchper-effort data from successiveSierra Leone/USSR research surveys. 43 2
Mauritania
15
O
IO
5
Longitude ("W) Figure 1 : Map of North Western Africa, showing the areas (A - H) used for aggregating the sea surfacetemperaturesextractedfromtheCOADS (courtesy ofDr. ClaudeRoy,ORSTOM).
;l 26
Figure 2. : Mean annual cycle of sea surfacetemperature off Sierra Leone (7"N to 9"N and 11"W to 14"W),derived from the COADS, 1950 to 1990(courtesyofDr.Claude Roy, ORSTOM).
J
F
M
A
M
J
J
A
S
O
N
D
Months
43 3
17
-
15
-
13
-
II
-
9-
Figure 3 :Tirne-latitude diagram of the mean seasonal cycle of sea surface temperature dong the Atlantic Coast of Africa from 23"Nto 7"N,based OR data extracted from theCOADS and averaged by month for the years 1950 to 1990 for the areas in Fig. 1. The seasonal distribution of Sardinella a~arifaoff Senegal, described in Garcia (1982) and Cury and Fontana (1 988), and off Sierra Leone, are superimposed.
434
I_
Figure 4 : Distribution of Sardinella maderensis in Sierra Leone waters, averaged over threemonths periods, from 1976 to 1990, basedon standardized catch-per-effort data from successive Sierra Leone/USSR research snrveys.
435
Figure 5 : Distribution of Sardinella aurita in Sierra Leone waters, averaged over three-months periods, from1976 to 1990, based on standardized catch-per-effort data from successive Sierra LeoneLJSSRresearch surveys.
436
CONTENT TABLE DES MATIERES Foreword ..............................................................................................................................................
P.7
Report of the Scientific meeting on Dynamics and Use of Sardinella Resources from Upwelling off Ghana andCôte d’Ivoire ................. P . 9 Annex 1: Assignments of DUSRU and Agenda ....................................p. 30 Annex 2: List of invited speakers and papers..........................................p. 33 Annex 3: Texts of the introductoryadresses............................................. p 35 Annex 4: List of participants .......................................................................................................... p 48
. .
.
.
AMAN A and FOFANA S Coastal Sea Surface Temperature as detected by the METEOSAT satellite and received at the Universityof Abidjan ..............p. 52
.
BAKUN A Global climate variations and potential impacts on the Gulf of Guinea SardinellaFishery .........................................................................
.
p 60
.
BARD F.X. Predation by tropical tunason Sardinella Possible induced changes in tunapopulations by increase of Sardinella abundance ........................................................................................................................... 85
.
BINET D Hypothesis accounting for the variability of Sardinella abundance inthe Northern Gulf of Guinea .............................................................p. 98
.
BORTEI-DOKU ARYEETEY E Kinsfolk and workers: Social aspect of labour relations among Ga-Dangme coastal fisherfolk ........p. 134
.
CHIKHI L., BONHOMME F and AGNESE J.F. The search for genetic differentiation of two sardine species (S aurita and S maderensis) .............p. 152
.
CURY P Comparison of the Ivoiro-Ghanaian fisheries dynamics with other upwelling system of the world .............................................................p. 169
.
. .
DURAND M H Mode of price formation for pelagic speciesand exploitation prospects in theless productive upwelling zone .................p. 194
.
PERRARIS S., KORANTENG K.A. Statistical analysis of canoe fishery data inGhana with particular reference to Sardinellas .........p. 205
.
437
.Small scale fishery for Sardinella in WCnin.....................p.223 . Ivsrian canoe fishery for Sardinella.............................................p. 232 hanaian fiishery for SardineIlas ...................p.243
.
.....
OFFI 9 Legal aspects in the management of the reso~rcesof small pelagics of the Ivoiro-Ghanaian ecosystem...............p 259
.
.
ORANTENG. K A and P.0. Tuna Baie boat baiting tirne and the availabilityof anchovy in Ghanaian waters...............p 263
.
.
The occur~eneeof zoo plankton off Tema dusing the p e r i ~ d1969 .1992................. .........................................................................p. 279
.Availability of Sardinella aurita eggs and larvae
g the period 1969.1992..................................................................
.
p 290
.
~.
ge determination and the hypothesis of annual double cycle of growth in ardinella aurita ....................................................................... 385
.
.
FIL1 9 Use of otoliths for stock discriminationand age estimation in ardinellas: Proposal for a regional research project in West Africa............. p 309
.
Ecslogical importance of the Hvsrian season .................................................................................................. 324
.
.
ROY C The CSte d’Ivoire and and changes ...............................................................................................*.................... 346
.
. p .362 9 A .Eeonomics of marine canoe fisheries in Ghana.........................p. 378
~A~~~ A and B SAMB Senegalese canoefishery for Sardinella............
.
.
TETTEY E West frica Markets for small pelagics .................................p. 395
.A. Sardinella market trends in the West African subregion ......................................................................................... $14
.
VAKILY M., PAULY D.(ICLARM) Seasonal movements of Sardinella off Sierra Leone.................................................................................................
43 8
426
Composition et Mise en page
B.P. 826 Cidex 1 Abidjan
Tél. : (225) 42 54 79
ORSTOM Éditions Dépôt légal : mai 1995 Impression Europe Media Duplication No3811
Diffusion 72, routed'Aulnay F-93143Bondy Cedex ISSN : 0767-2896 ISBN : 2-7099-1258-9
Photo de couverture : Débarquement de sardiilelles mildes(S. aurita) au port d'Abidjan, 1993. Clichd :Isabelle Bard 