jjg^HT OF Cq^ NOAA Technical Memorandum NMFS-SEFSC - 370 Historical Document: Life History and Fisheries of Atlantic Bluefin Tuna DATA LIBRARY WOODS HOLE CXZEAN'CGRAPHIC INSTTTLmCJI^ Frank J. Mather, III Woods Hole Oceanographic Institution Woods Hole, Massachusetts 02543 John M. Mason, Jr. New York Department of Environmental Conservation Division of Marine Resources East Setauket, New York 1 1733 Albert C. Jones U.S. Department of Commerce, NOAA National Marine Fisheries Service Southeast Fisheries Science Center 75 Virginia Beach Drive Miami, Florida 33 149 U.S. Department of Commerce National Oceanic and Atmospheric Administration C\ L, National Marine Fisheries Service / -3^ Southeast Fisheries Science Center f- 75 Virginia Beach Drive '53b Miami, Florida 33 149 1 ^^6 c 3 .^qWMOS/^,,, '^Me^fT of NOAA Technical Memorandum NMFS-SEFSC-370 DATA LSBHAHv WOODS HOLE OCEAK'CXSRAPHIC INSTrrUTIW Historical Document: Life History and Fisheries of Atlantic Bluefin Tuna Frank J. Mather, III Woods Hole Oceanographic Institution Woods Hole, Massachusetts 02543 John M. Mason, Jr.' New York Department of Environmental Conservation Division of Marine Resources East Setauket, New York 17733 Albert C. Jones U.S. Department of Commerce, NOAA National Marine Fisheries Service Southeast Fisheries Science Center 75 Virginia Beach Drive Miami, Florida 33149 U.S. Department of Commerce Ronald H. Brown, Secretary National Oceanic and Atmospheric Administration D. James Baker, Undersecretary for Oceans and Atmosphere National Marine Fisheries Service RoUand A. Schmitten, Assistant Administrator for Fisheries ' Formerly with Woods Hole Oceanographic Institution. This technical memorandum is used for documentation and timely communication of preliminary results, interim reports, or special purpose information, and has not undergone external scientific review. Julv 1995 NOTICE The National Marine Fisheries Service (NMFS) does not approve, recommend, or endorse any proprietary product or material mentioned in this publication. No reference shall be made to the NMFS, or to this publication furnished by NMFS, in any advertising or sales promotion which would indicate or imply that NMFS approves, recommends, or endorses any proprietary product or proprietary material mentioned herein or which has as its purpose any intent to cause directly or indirectly the advertised product to be used or purchased because of NMFS publication. This report should be cited as follows: Mather, F.J., J.M. Mason, and A.C. Jones. 1995. Historical document: life history and fisheries of Atlantic bluefm tuna. NOAA Technical Memorandum NMFS-SEFSC - 370; 165 pp. Woods Hole Oceanographic Institution Contribution Number - 8986 AUTHORS' NOTE AND ACKNOWLEDGMENTS This report is based on extensive studies of Atlantic blucfin tuna carried out over many years by the senior author. The report was prepared in the late 1970s but not published. It is published now to provide a historical review of the biology and fisheries of bluefin tuna, based on knowledge available at that time. More recent studies have expanded the information base for bluefm tuna; however, that information is available, mostly in reports of the International Commission for the Conservation of Atlantic Tunas (ICCAT). Consequently, we have not attempted to include a review of that information in this document. We acknowledge financial and other support received from the following institutions and persons: the National Science Foundation, the National Marine Fisheries Service, the Sport Fishing Institute (now the American Sportfishing Association), the National Sea Grant Program, the Charles W. Brown, Jr. Foundation, the Lou Marron Science Fund, Mrs. Raymond J. Kunkel, Mr. William K. Carpenter, the Lemer Marine Laboratory, the Woods Hole Oceanographic Institution, and the Associates of the Woods Hole Oceanographic Institution. In addition to persons mentioned here and in the text, numerous individuals and fishery and research organizations contributed valuable information. Copies may be obtained by writing: National Marine Fisheries Service Southeast Fisheries Science Center 75 Virginia Beach Drive Miami, Florida 33 149 or National Technical Information Service 5258 Port Royal Road Springfield, Virginia 22161 (703)487-4650 FAX; (703) 321-8547 Rush Orders: (800) 336-4700 CONTENTS I. INTRODUCTION ' II. METHODS AND DEFINITIONS 2 A. Methods and Materials 2 B. Definitions 2 1. Bluefin Tuna 2 a. Length - b. Weight 3 2. Size-Age Groups 3 3. Trap Fishery Terminology 3 III. AGE AND GROWTH 5 A. Introduction -^ 8. Methods of Determining Age and Growth and Their Application to Bluefin Tuna 5 1. Size Frequencies ^ 2. Counts of Marks on Hard Parts 5 a. Scales ^ b. Vertebrae 6 c. Otoliths ^ 3. Release-Recapture Data for Tagged Fish 8 C. Results ^ 1. Sizes of Fish at Determinable Ages 8 2. Growth of Bluefin Tuna in Its First Year of Lift 9 3. Growth of Young Bluefin Tuna 1 1 4. Seasonal Changes in the Length- Weight Relationship of Large Bluefin Tuna 12 5. Differential Growth of Males and Females 13 6. Ultimate Length '3 7. Discussion '^ IV. DISTRIBUTION AND FISHERIES 15 A. General Distribution '5 B. Development of the Major Fisheries 15 C. Distribution and Fisheries by Areas 18 1. Introduction '^ 2. Western Atlantic '8 a. Newfoundland to Cape Cod 1 8 i. Newfoundland '8 ii. Gulf of St. Lawrence 19 iii. Nova Scotia '9 iv. Cape Cod to Maine 22 b. Cape Hatteras - Cape Cod (Coastal Waters) 25 c. Atlantic Ocean Outside 200 Meter Contour N of 35°N and W of 40''W 27 d. Bahamas and Southeastern Florida 27 e. Gulf of Mexico and Caribbean Sea 29 3. Atlantic Oceanic Waters 29 4. Eastern Atlantic Islands 34 a. Azores 34 b. Madeira 35 c. Canary Islands 35 5. Eastern Atlantic 36 a. Northeastern Atlantic 37 b. Norway -'' c. North Sea 38 d. Bay of Biscay 39 e. West Coast of Portugal 42 f West Coast of Morocco 42 g, I bero- Moroccan Bay Trap Fisheries 42 h. Trends in Eastern Atlantic Fisheries 46 6. Mediterranean and Black Seas 46 a. Western Mediterranean 48 b. Central Mediterranean 53 c. Eastern Mediterranean and Black Sea 63 d. Summary ^^ V. SPAWNING 68 A. Introduction ^° B. Definitions 68 C. Criteria for Determining Seasons and Areas of Spawning, and Their Limitations 68 1. Presence of Ripe Adults 68 2. Presence of Pelagic Eggs and Larvae 69 D. Spawning Areas and Seasons 70 1 . Mediterranean and Black Seas 70 a. General Information 70 b. Specific Occurrences 70 2. Eastern and Central North Atlantic and the South Atlantic 72 a. General Information 72 b. Specific Occurrences 73 3. Western North Atlantic 80 a. General Information 80 b. Specific Occurrences 80 E. Relationships Between Environmental Factors and Spawning Behavior 84 1 . Introduction ^4 2. Mediterranean Sea 86 3. Eastern Atlantic 87 a. Distribution, Objectives and Nature of Research 87 b. Portugal 88 c. Spain 88 d. Morocco 90 e. Other Eastern Atlantic Areas 91 4. Western Atlantic 91 a. Introduction 91 b. Gulf of Mexico 91 c. Straits of Florida 91 d. East and North of the Bahamas 91 e. Northeastern United States 92 F. Feeding Activity During the Spawning Season ... 92 G. The Spawning Act 92 H. Maturity and Fecundity 92 1. Age and Size at First Maturity: 92 2. Fecundity 93 3. Maximum Size at which Bluefin Tuna Spawn 94 I. Discussion and Conclusions 94 1. Mediterranean and Black Seas 94 2. Eastern Atlantic 95 3. Western North Atlantic 95 4. Overall Situation 95 VI. MIGRATIONS AND STOCK IDENTIFICATION 96 A. Introduction 96 B. Methods and Materials, and Definitions 96 1. Methods and Materials 96 a. Deductive Methods 96 b. Methods Used Mainly for Identifying Stocks 96 c. Methods Used Mainly for Studying Migrations 96 d. Development of Bluefin Tuna Tagging 96 e. Summary of Bluefin Tuna Tagging Programs 97 2. Definitions 97 3. Hypothetical Migration Model 97 C. Studies of Migrations and Stock Identity 98 1 . Mediterranean and Eastern Atlantic 98 a. Introduction 98 b. Migrations Between the Mediterranean and the Eastern Atlantic 99 c. Migrations and Stocks within the Mediterranean 104 d. Migrations and Stocks in the Eastern Atlantic 108 i. Large and Medium Fish 108 ii. Small Fish 1 1 1 iii. Very Small Fish 1 12 2. Western Atlantic 1 13 a. Introduction 113 b. Large Fish 113 c. Medium Fish 124 d. Small Fish 127 e. Very Small Fish 129 3. Trans- Atlantic and Trans-Equatorial Migrations, and Atlantic Stocks 129 a. Introduction 129 b. Trans-Atlantic Migrations 129 i. Giant Fish 129 ii. Medium Fish 130 iii. Small fish 130 c. Trans-Equatorial Migrations 132 d. Stock Identification 132 i. Biometric Studies 132 ii. Biochemical Studies 132 iii. Conclusions 132 VIL DISCUSSION, RECOMMENDATIONS, AND CONCLUSIONS 133 A. Introduction 133 B. Age and Growth 133 C. Distribution, Migrations, and Spawning 133 1. Introduction 133 a Very Small Fish 133 b. Small Fish 133 i. Immature Individuals 133 ii. Mature Individuals 133 c. Medium Fish 133 d. Large Fish 133 2. Western Atlantic 134 a. Very Small Fish 134 b. Small Fish 134 c. Medium Fish 134 d. Large Fish 135 3. Eastern Atlantic and Mediterranean 135 a. Very Small Fish 135 b. Small Fish 136 c. Medium Fish 137 d. Large Fish 137 4. Comparison 139 D. Identification of Stocks 140 E. Recommendations 143 VIII. BIBLIOGRAPHY 145 A. Introductory Notes 145 B. References 145 IV I. INTRODUCTION The objective of this worlc is to review and summarize available infor- mation on the fisheries, distribution, and other aspects of the Hfe history of the Atlantic bluefm tuna, Thunmts thynnus thyimus (Linnaeus 1 758) (Fig- ure 1). The need for such a review is emphasized by the concern over the staUis of the Atlantic bluefin tuna stocks and the recent enactment of measures for the conservation of this species by the International Commission for the Conservation of Atlantic Tunas (ICCAT). The wastefulness of harvesting Atlantic bluefin tuna at extremely small sizes was recognized intuitively by d' Amico ( 1 8 1 6, cited by Pavesi 1 887). Sicilian laws of 1 796 and 1 80 1 prohib- ited the catching of small bluefin tuna (Avolio 1805 in Parona 1919). This restriction was apparently dropped dur- ing revisions of Italian fishery regula- tions in 1877 and 1882 (Pavesi 1887), but a minimum size limit of 60 cm was imposed in 1927 (Mussolini and Belluzzo 1927). In the 1960s, a mini- mum size limit of 90 cm was in effect in Italy (Sara 1968, Miyake 1976, Maldura 1965). The effect of harvesting bluetln tuna at small sizes was estimated theo- retically by Shingu et al. (1975) and empirically by Mather (1974). Both works indicated that the capture of one thousand tons of young bluefin tuna precluded the subsequent capture of many thousand tons of larger individu- als. Action in 1 975 by ICCAT in regu- lating the bluefin tuna fisheries finally recognized in principle the need to man- age the fisheries for this economically important species. The regulations which became effective August 10, 1975, are as follows (Miyake 1975): First — That the contracting par- ties take the necessary measures to pro- hibit any taking and landing of bluefin tuna (Thunnus thynnus thynnus) weigh- ing less than 6.4 kg. Notwithstanding the above regulation, the contracting parties may grant tolerances to boats which have incidentally captured blue- fin tuna weighing less than 6.4 kg, with the condition that this incidental catch should not exceed 15% of the number offish per landing of the total bluefin tuna catch of said boats or its equiva- lent in percentage by weight. Second — That as a preliminary step, the contracting parties that are actively fishing for bluefin tuna ( Thunnus thynnus thynnus) or those that ment (National Marine Fisheries Ser- vice, 1975, 1976) Canada, France, and Japan have also put the ICCAT regula- tions into effect (Caddy and Burnett 1976, Kume 1976, 1977). Whether these regu lations were en- acted in time, and are adequate to re- store the Atlantic bluefin tuna fisheries to their potential and former impor- tance, is questionable. It is certain, hovy- ever, that better knowledge of the life history of the bluefin tuna is a prerequi- site to effective management of its fish- eries. By general consensus, knowl- Figure 1. Atlantic bluefin tuna {Thunnus thynnus thynnus). incidentally catch it in significant quan- tities shall take the necessary measures to limit the fishing mortality of bluefin tuna to recent levels for a period of one year. At the 1975 meetings of the ICCAT Commission, this second item was extended for an additional two years, with provision for its review at the 1976 meetings of the ICCAT Coun- cil. The member nations are respon- sible for enforcement of ICCAT regula- tions within their territorial waters. For the United States, the Atlantic Tunas Convention Act of 1975 empowered the Secretary of Commerce to imple- ment regulations established by ICCAT. The National Marine Fisheries Service (NMFS) drew up a set of regulations, and is responsible for their enforce- edge of stock identity and migratory patterns is of prime importance. The aspects of the fisheries and life history herein discussed include: age and growth information; catch sta- tistics; size, sex, and age composition of the landings in various areas; spawn- ing and development; migration; and stock identification. Environmental pa- rameters and their possible influences on the distribution, migrations and spawning of the species are also con- sidered. Hypotheses on the migrations of bluefin tuna and stock identity are discussed. Finally, we state our own conclusions in regard to these matters and make some recommendations for future research. 1 II. METHODS AND DEFINITIONS A. METHODS AND MATERIALS Personal research by the authors during 1950-1976 has produced much of the knowledge which is summarized herein. Extensive additional informa- tion was obtained from the literature, by participation in meetings, and through correspondence and conversa- tions with scientists, fishery experts, and fishermen. The authors and their colleagues obtained a great mass of data and infor- mation on the species and its fisheries directly. They examined the landings and often observed or participated in the operations of various commercial and recreational fisheries over a great geographical range. They also partici- pated in many exploratory fishing cruises of U.S. and foreign research vessels. The Cooperative Game Fish Tag- ging Program of the Woods Hole Oceanographic Institution (a joint pro- gram with the National Marine Fisher- ies Service since 1973), initiated by the senior author in 1954, has provided much of the information on migrations and mortality rates of Atlantic bluefm tuna. Exchange of information and co- operation with marking programs of other nations has been extensive. Also, much information has been obtained through participation in meet- ings of a variety of groups. These in- clude international regulatory agencies and advisory groups, such as ICCAT, the Panel of Experts for the Facilitation of Tuna Research of the Food and Ag- riculture Organization (FAO) of the United Nations, and national, regional and state fishery agencies. We have also attended numerous meetings of non-governmental research and fish- ery groups or associations. Further knowledge was obtained by correspondence and conversations with representatives of agencies of the types mentioned above, and with indi- viduals concerned with the Atlantic bluefm tuna in many areas and several nations. An important additional source of information was a thorough search of the literature. This was most intensive B. DEFINITIONS 1. Bluefm Tuna The subject of this paper is the Atlantic bluefin tuna, Thunnus thynnus thynnus, as distinct from the southern Fork Length Figure 2. Fork length measurement. at the initiation of our studies and has been continued since then to the extent that time has permitted. This experi- ence has enabled us to locate many important and little known references. Our extensive contacts and data sources, increasing over the years, have enabled us to maintain close connec- tion with the entire Atlantic bluefin tuna situation. In the course of our investigations it became increasingly apparent that this species cannot be studied success- fully on a piecemeal or limited area basis. Our approach, therefore, has been an attempt to observe and describe the species and its fisheries over its entire geographic range and during each stage of its life cycle. Our methodology has been a combination of original research and a review of the findings of other investigators in all the nations con- cerned with Atlantic bluefin tuna. bluefin tuna, Thunnus maccoyii (Castelnau 1872) and the Pacific blue- fin tuna, Thunnus thynnus orientalis (Temminck and Schlegel 1844). The ranges of the Atlantic and the southern bluefins overlap off South Africa and in the South Atlantic, whereas those of the Atlantic and the Pacific bluefins are apparently separate (Gibbs and Collette 1 967, Talbot and Penrith 1 968, Fisher- ies agency of Japan 1974, 1975, 1976, 1977). Throughout this report, the terms Atlantic bluefin tuna, bluefin tuna, blue- fin, tuna, fish, and individual refer to Thunnus thynnus thynnus unless other- wise identified. a. Length The most widely used measure- ment of length for bluefin tuna over 1 2 cm is fork length (FL). This is the straight line length, measured by cali- pers or equivalent instruments, from Table I. Size-age groups used to discuss much of the biological and fisheries data for Atlantic bluef'in tuna in this report. Group Length Weight Approximate Age (cm) (in) (kg) (lbs) (years) Very small < 50 <20 Small 50-120 20-48 Medium 120-185 48-76 Large or Giant >185 >76 <2.5 <5 2.5-32 5-70 1 -4 32 - 122 70 - 270 5-8 >122 >270 >8 the snout (tip of upper jaw) to the pos- terior medial margin of the caudal fin (Figure 2). In this paper, length mea- surements exceeding 12 cm are fork lengths unless defined otherwise. An alternative method of measuring fork length is with a tape and following the lateral contour of the body. Lengths measured by this method have been referred to as tape lengths, curved lengths or flank lengths. The slope of a linear regression which Schuck and Mather fitted by inspection to a plot of straight lengths against curved lengths for each of a series of western north Atlantic bluefin tuna from 35 to 270 cm longindicated that the straight length was 0.958 times the curved length. Sev- eral other authors have published con- version factors for these parameters. Of course this relationship would vary slightly with the length-girth (or length- weight) ratio of the fish. The lengths of juvenile specimens less than 1 2 cm long and of larvae are usually measured in standard length (SL). This is the length from the snout to the end of the vertebral column. In this paper, length measurements of blue- fin tuna less than 12 cm long are stan- dard lengths unless defined otherwise. b. Weight Weight data for fish are presented in this paper in terms of round, whole, or live weight (the weight of the entire fish) unless othenvise stated. In some fisheries, it is necessary to collect weight data for fish in the condition in which they are sold. This may be in several forms depending on the extent to wh ich the fish have been butchered. Conver- sion factors for obtaining round weights from weights in the various other con- ditions are available in the literature and at the Woods Hole Oceanographic Institution. Weights of individual fish and av- erage weights of fish are recorded in kilograms, unless otherwise stated. Weights of landings and catches (of numerous fish) are recorded in metric tons unless otherwise stated. 2. Size- Age Groups Much of the biological and fisher- ies data for the Atlantic bluefin tuna will be discussed in terms of the size- age groups shown in Table 1 . The lim- its of these groups were selected to correspond as closely as possible to sizes and estimated ages at which the migratory and distributional patterns of the species undergo distinct changes (Mather 1 964b). The ages for the re- spective sizes are from Mather and Schuck ( 1 960). These size groups were set up on the basis of data from the summer fisheries in the northwestern Atlantic and the spring fishery off the Bahamas. They appear, however, to be reasonably applicable to bluefin in the eastern Atlantic and the Mediterranean Sea. 3. Trap Fishery Terminology Specialized terms have been used to describe the bluefin tuna trap fisher- ies of the eastern Atlantic and the Medi- terranean Sea. Since these terms are often misunderstood or poorly trans- lated in the literature, an explanation of them seems desirable. The true bluefin tuna traps are very large, complex installations set at spe- cial locations to harvest runs (periodic migratory passages) related to the spawning of these fish. "Arrival" fish are essentially fat, maturing individu- als, generally traveling eastward in late April, May, and June. "Return" fish are essentially lean, spent individuals, gen- Table 2. A glossarN' of some of the more important tuna fishery terms in English. French, Italian, Portuguese, and Spanish. English French Italian Portuguese Spanish Irue Tuna Trap Madrague Thonaire' Tonnara' Minor Tuna Trap Petite Madrague Tonnarella Thonaire Mixle Arrival Arrive Course Corsa Retum Rctour Ritomo Genetic Gcnetique Genctico Erratic Erratiquc l>ratico \imacoa Almadraba Almadrabilla Direito Derecho Arribada Revcz Reves Genetico Erratico ' The words thonaire and tonnara arc also used locally for oilier gears which catch tunas or tuna-like fishes erally traveling westward in July, Au- gust, and early September. Some of the younger spawners, which have not yet discharged their eggs, are sometimes taken with the return fish. The arrival fish are not only more robust than the return ones, but their flesh is fattier and usually commands a higher price. In most areas, traps were fished for only one run — either as arrival or return traps. However, along the south- em Atlantic coasts of Portugal and Spain, many traps were altered to fish for each season, and in this way they fished both runs. Smaller and less important traps take bluefin tuna throughout much of the year, but these fish are mainly im- mature individuals or larger ones which have completed their spawning cycle and are more widely dispersed, pre- sumably in search of food. Roule (1914a, 1914b) proposed the terms "genetic" for maturing blue- fin tuna, and "erratic" for those which had spawned and whose behavior was dominated by search for food. These terms occur in many works on these fisheries. A brief glossary of some of the more important terms, in English, French, Italian, Portuguese, and Span- ish is listed in Table 2. III. AGE AND GROWTH A. INTRODUCTION Knowledge of the age-size rela- tionship is not only important from the biological viewpoint but is also essential for effective management of a fishery. Until recently, there has been general agreement on the sizes of Atlantic bluefin tuna at ages up to 12-14. Preliminary results for ages up to 26 have been presented in 1975 and 1976, but these are controver- sial. B. METHODS OF DETERMINING AGE AND GROWTH AND THEIR APPLICATION TO BLUEFIN TUNA 1. Size Frequencies Age groups show up as modes when the sizes of sufficient numbers of bluefin tuna are plotted. These modes are usually distinct for small individuals (up to 125-150 cm long), but become less distinguishable with larger fish. The length-weight ratio of large bluefin fluctuates greatly with the seasons (Tiews 1963). Therefore length measurements are more suit- able for size frequency analyses than weights. Growth may be estimated fi-om the progression of modes in plots of size data for consecutive time intervals. Size-frequencies were evidently used, although the methodology was not described, by d'Amico (1816, in Heldt 1930) and Bourge (1908, in Roule 1917), in estimating the sizes of juvenile Mediterranean bluefin tuna in their first four months of life. Piccinetti and Piccinetti-Manfrin ( 1 970) presented a more detailed and precise study of their growth through this period, describing the methods they had used. Westman and Gilbert ( 1 94 1 ) and Westman and Neville (1942) were evidently the first to study the growth of larger bluefin tuna by this method. They traced the growth of young in- dividuals taken off Long Island, New York, in the summers of 1938 and 1941 by analyzing their length fre- quencies. They established the age in years of each size group by counting annuli on scales. Buser-Lahaye and Doumenge (1954) and Doumenge and Lahaye (1958) likewise analyzed length fre- quencies to estimate the ages of small bluefin tuna caught off the Mediter- ranean coast of France during 1953 and 1954. They used Sella's (1929a) data, however, to establish the esti- mated age of each size group. More extensive studies of the summer growth of western Atlantic bluefin tuna, using counts of annuli on hard parts as well as size frequen- cies, were presented by Mather and Schuck(l960). Furnestin and Dardignac (1962) were the first authors to trace the growth of 7". ihynnus ihynniis through most of the first two years of its life. They used size frequencies of young fish taken off the Atlantic coast of Morocco, where, after attaining a length of about 32 cm, they are avail- able throughout the year. Tiews (I960) used the frequen- cies of eye diameters, which he as- sumed to be related to the age of the fish, as well as to its length, to esti- mate the ages of large bluefin tuna caught in the North Sea in the sum- mer and fall of I95<). 2. Counts of Marks on Hard Parts Ages of bluefin tuna have been estimated by counting marks, usu- ally called annuli. which were thought to have been laid down annually, on certain hard parts of the fish. The parts used included scales, vertebrae, and otoliths (Figure 3). The relative merits of these meth- ods were extensively discussed at the "conference d' experts pour I'examen des melhodes scientifiques et tech- niques ^ appliquer a I'etude des poissons de la famille des Thonides" (hereafter referred to as "Conference of experts") held at Madrid and Cadiz, Spain, 16-22 May 1932 (Anonymous 1932b), and by others as cited here. a. Scales Age determinations from scales are based on counting the "checks" or areas where two or more circuli are close together, instead of being more widely and evenly spaced as they are on most of the scale's sur- face. Corson (1923a, 1923b) presented the first age determinations for Atlan- tic bluefin tuna of which we have knowledge. These were based on scales taken from a small number of young fish caught off Long Island, New York, in September 1923. He used scales from the posterior part of the fish, having found those from the shoulder to be unreadable because of streaks and globules of oil within them. At the "Conference of experts" (Anonymous 1 932b), F. de Buen and Sella said that scales did not furnish Figure 3. Bluefin tuna vertebra showinsi two annuli. interesting information on the age of tiie biuefin tuna, but Frade showed that growth lines were clearly indi- cated on scales of young individuals. The experts concluded that scales were not useful for determining growth in this species, except for young individuals. Spagnolio (1938) estimated ages of from 3 to 6 years from scales of biuefin tuna caught in traps off south- em Italy and northeastern Sicily. She presented illustrations of these scales but did not report the sizes of the fish. She concluded that readings from scales should be checked against readings from vertebrae of the same individual. She maintained that if the method were validated, scales would be preferable to vertebrae for deter- mining ages of small biuefin tuna, up to 40-58 kg or 5-6 years of age, since scales were easier to collect and to examine than vertebrae. Spagnolio used scales from the caudal and lat- eral parts of the body, rejecting those from the corselet because they were too thick and opaque. The scales which she used were preserved in formalin, and later were soaked in water or an alcohol solution or glyc- erine before examination. She found no marked advantage in using either of the last two fluids instead of wa- ter. Westman and G ilbert ( 1 94 1 ) and Westman and Neville (1942) used readings of annuli on scales, as well as analyses of length frequencies, to determine ages of young (1-7 year old) biuefin tuna taken off Long Is- land, New York, during the summers of 1938 and 1941. These authors also offered tentative age determinations from scales for a few much larger biuefin tuna, up to 250 cm and 275 kg, with an estimated age of about 1 8 years. They used carefully selected thin and round scales from the side of the body, just below the lateral line in the area below the base of the second dorsal fin. Projected impres- sions of the scales facilitated their readings of those with more than five annuli. Mather and Schuck (1960) esti- mated ages of 0-4 years from scales of biuefin tuna taken off the north- eastern United States during several summers. They relied more heavily on length frequencies for studying the growth of young individuals, however, and on vertebral annuli for the older ones. These authors used scales from the general part of the tuna's body where Westman and Gil- bert ( 1 94 1 ) and Westman and Neville (1942) collected theirs. They made impressions of the scales on cellu- loid (Arnold 1 95 1 ), and counted an- nuli on magnified projections of these impressions. None of the above authors used the large and thick scales of the corse- let for age determinations, but F.S. Russell mentioned (personal commu- nication) that he had found a pos- sible method of determining ages from these scales. He found that they were built up of lamina, like the pages in a book, which might represent years of growth. These lamina could be separated after the scales had been soaked in a weak solution of acetic acid. The most important use of scale readings in aging biuefin tuna has been in validating determinations made by other methods (Westman and Neville 1942, Mather and Schuck 1960). b. Vertebrae F. de Buen (1925) estimated, by counting its vertebral rings, that a male biuefin tuna 206 cm long (from snout to tips of caudal) and weighing 1 19 kg was 12 years old. This fish was caught July 4, 1923. in the Barbate trap near Cadiz, Spain. Sella (1929a) presented mean lengths and weights of Mediterranean biuefin tuna for ages 1-14, as esti- mated by counting the rings in the centra of vertebrae. This was the first study which described the growth of biuefin tuna through most of the size range ordinarily encountered At the "Conference of experts" (Anonymous 1932b), Sella and the other experts discussed the use of vertebrae forage determination. Sev- eral methods of preparing vertebrae for examination were described, and it was noted that they could be exam- ined without special preparation. He described three instruments which he used in the study of vertebrae and stated that the annuli were generally better defined in vertebrae from the caudal trunk. Heldt reported that, in each of five biuefin tuna of different sizes which he had examined, the number of annuli on each of the ver- tebrae was the same. He also noted that double rings, which should not be counted as two years, sometimes occurred. Sella and F. de Buen main- tained that the vertebral rings repre- sented years of age, but the latter noted that, because of the time of spawning, the first ring did not cor- respond exactly to one year. Frade reported finding 16 vertebral annuli for a 263 cm fish, which exceeded the maximum age reported by Sella (1929a). Several workers reported on the growth of biuefin tuna from different areas as determined from readings of vertebral annuli in the period 1956- 1962. These include Hamre (1958, i960) (Norway), Rodriguez-Roda (1960) (Spain), Vilela and Pinto ( 1 958), Vilela ( 1 960), and Frade and Vilela (1962) (Portugal), and Mather and Schuck (1960) (northeastern United States). Mather and Schuck used the technique of Galtsoff ( 1 952) to stain many of the vertebrae which they used for age determinations. Rodriguez-Roda (1964a) presented thorough mathematical interpreta- tions of his I960 results. Butler ( 1 97 1 , 1975) counted the rings in the vertebrae of large biuefin tuna taken in Canadian waters in 1 966 and in 1974 and estimated their ages as 1 1 to 22 or more years. Myklevoll used this method to determine the ages of biuefin tuna taken in Norwe- gian waters in 1974 (Caddy and But- ler 1976). These estimates ranged from 1 4 to 2 1 years. Butler and Myklevoll were the first investigators to estimate the ages of significant numbers of bluefm tuna more than 14 years old. Berry et al. (1977) discussed in great detail the techniques of aging bluefin tuna by reading annuli on vertebrae and otoliths. They de- scribed methods of storing, staining and examining vertebrae, and their interpretation of the markings on them. They found that immediate freezing and freezer storage produced better results than the other methods of preservation which they tested. Im- mediate staining and reading, how- ever, was probably the most satisfac- tory procedure. They noted that stain- ing time varied with the size of fish, and that the inner rings, particularly in large fish, stained before the outer ones. Their interpretation of marks on vertebrae was complex. It involved ridges, grooves, fimbriated lines, and stained and unstained rings. They noted that several stained rings might occur early in the staining process within an annular zone, and that these might coalesce in various ways as staining progressed. Berry et al. (1977) did not in- clude size-for-age data or age data for individual fish, but they presented extreme and average vertebral ages, determined from readings of verte- bral annuli, for bluefin tuna in vari- ous weight ranges. They tentatively recommended the use of vertebrae for estimating ages up to about 10 years for this species. From 1932 until 1974, counting vertebral annuli was generally re- garded as the most satisfactory method of aging large Atlantic blue- fin tuna. c. Otoliths The use of otoliths in determin- ing ages of T. Ihvnnus ihynnus was first investigated in the 1920s, bul important results did not appear until 1975. Despite the fact that F. de Buen (1925) had been discouraged by the difficulties encountered in extracting otoliths, Frade ( 1 925) described pro- cedures for their relatively rapid re- moval, and the nature of the growth zones on them. These descriptions, accompanied by excellent illustra- tions, suggested that this was a prac- tical method of age determination. At the "Conference of experts" (Anonymous 1 932b), however, Frade reported that, although he had found zones of growth in thin sections of otoliths, he was not sure that these sections included all of the years of growth. He attributed his uncertainty to the irregular shape of the otoliths. The experts concluded that growth zones were laid down on otoliths, but that it was difficult to assign absolute ages by counting these zones. They therefore recommended that this tech- nique be used mainly to confirm age determinations obtained by other means. Frade and Vilela ( 1 962) referred to age determinations from otoliths by Frade (1950), but we have not seen the latter work. No further ref- erences to Frade's early work with otoliths have come to our attention. The next attempt to determine ages of bluefin tuna from otoliths was by Nichy (Nichy and Berry 1 976). They developed techniques for estimating the ages of bluefin tuna similar to those of Frade ( 1 925). They used otoliths from large individuals caught off Prince Edward Island in the Gulf of St. Lawrence, Canada, in 1974. Caddy and Butler (1976) clas- sified large bluefin tuna taken in Ca- nadian waters in 1974 and 1975 by yearclasses, using Nichy and Berry's techniques and some of their deter- minations in addition to their own. They did not report the sizes of these fish. Caddy et al. (1976) present the most complete study available on the growth of bluefin tuna as determined from otoliths. They used this method to determine the ages of large indi- viduals (age 10 and over) caught in Canadian waters during the summer and fall of 1975. They calculated pa- rameters for von Bertalanffy growth equations for males and females, us- ing their data for fish of ages 10 or greater and Mather and Schuck's (1960) results for fish of ages 1-4. Their work was the first to indi- cate different growth rates for the two sexes. Butler et al. ( 1 977) repro- duced the above material with the addition of preliminary age data from the otoliths of 60 large bluefin taken in Canadian waters in 1976, and pre- sented some modifications of the Nichy and Berry (1976) and Caddy and Butler ( 1 976) techniques. Berry et al. ( 1 977) discussed the removal, storage, measuring, mount- ing and sectioning of otoliths. They described the sections and their in- terpretation of the markings on them. They found that, at ages greater than 10 or II, hyaline bands occurred in pairs which represented a single an- nular zone. They discussed definitions of annuli on vertebrae and otoliths, and pro- posed the following tripartite hypoth- esis: a) Major discernible markings on ver- tebrae and otoliths of Atlantic blue- fin tuna do nol have to be pre- sumed to be annuli. b) Within each year of life, multiple markings are successively formed on vertebrae and in otoliths. These multiple markings may appear in prepared specimens of vertebrae and otoliths as irregular combina- tions Single markings that repre- sent the end of a year of life (an- nuli) may be distorted by varia- tions of the within year multiple markings. c) Annuli may he deciphered by con- sidering the nature in which they form and by interpreting the varia- tions that may exist within and be- tween individual prepared samples. Berry et al. ( 1 977) did not present age data based on readings of annuli on otoliths. 3. Release-Recapture Data for Tagged Fish When size data at release and recapture of a given fish are avail- able, its growth during the period at liberty is established. If age at the time of release can be determined, age at the time of recapture can like- wise be calculated. Positive age and growth information thus obtained can provide a valuable check on estimates based on indirect methods (Mather 1980). C. RESULTS 1. Sizes of Fish at Determinable Ages The first objective of most age and growth studies is to determine the mean sizes offish at each year of age for which this is possible. Ideally the sizes should be determined at one year intervals after the date of spawn- ing. This was done by Sella (1929a) and Rodriguez-Roda (1960, 1964a), but most other investigators have been limited to seasons in which suf- ficient material was available. Sella (1929a) presented the first important study of the age and growth of bluefin tuna. This was based on counts of annuli on veilebrae from more than 1,500 bluefin tuna caught in June, during the spawning season. The individuals whose ages were as- sessed as 3 or more years had been taken off Tripolitania (western Libya). Sella noted that these results did not differ noticeably from those for bluefin tuna taken in other parts of the Mediterranean. He assessed ages of 1 and 2 years for individuals which had been captured in the Adriatic Sea, since fish of these sizes were not available off Libya. He pre- sented his results in terms of mean length and mean weight for each year of age from 1 through 14. Westman and Gilbert (1 94 1 ) and Westman and Neville (1942) pro- duced the first important informa- tion on the growth of bluefin tuna taken in the western Atlantic. Their works provided well supported LnielhtKenb7S«lb(192»a) Wdghl given b;SeUi(l»2»a) • LrngtbglvMib; WMta>aniDdCUb»r1(l»4l) O Lcagtbtlv«b7Hamn(l9C0) X Uopb given b7VUd*«taL(l9tO) 2S0-, 2*0- 220 200- 190- 160- liO 120- m- eo- 60- E % O L*iiglligtvenb]>TIECOO MMTLANO-CAROLINAS yiAMr AHEA GULFOFWXICO I tlWiItt |W>TtytH| gitfam I Hovtme* I occiiittn I j«tiu«>r' 40p z so; 2S 1 z 20 u 15 ' 10 Figure 7. Lengths of bluefin tuna less than 50 cm long (young of the year) collected in the western Atlantic and the Gulf of Mexico. The curve of estimated growth was fitted by inspection (Mather and Schuck 1 960). western Atlantic fish are about 50 cm long, according to Fumestin and Dardignac (1962) and Mather and Schuck (1960). Larger sizes are re- ported for 1 -year-old tuna by Rodriguez-Roda (1960) for eastern Atlantic fish (55 cm) and Sella ( 1 929a) for specimens from the Medi- terranean Sea (64 cm). 3. Growth of Young Bluefin Tuna Westman and Seville (1942) were the first to show that young bluefin tuna grow much more rap- idly during the summer than during the winter. They reached this conclu- sion by tracing the growth of young fish taken off Long Island, New York, through the period July I -October 16, 1 94 1 , from length frequency data. Mather and Schuck ( 1 960) stud- ied this matter in greater detail. Their data included some specimens taken during both summer and winter in their first year of life (age 0), but data were available for ages 1-4 for the period July-October 16 only. The growth rates in the latter period were very rapid (about 3.8 cm per month), and it was evident that their average II 5 u I ■ •. o 80 - /^^ .A 70 - A,K ■ Aao-^^* " 60 " r* ' ■ ?A°* 50 _ ?aA ^ ^,^* YEAR CLASSES ol-^ 1949 40 - /' 1950 / 195) / f 1952 30 J 1 — 1 — 1 — 1 1 — 1 — 1 — 1 — 1 1 — 1 — I — 1 — 1 — 1 — 1 — 1- 1953 _j — 1 — 1 — 1 — 1 1 — 1 IX XI VII IX XI MONTHS III VII IX Figure 9. Curve of linear growth of young bluefin tuna (caught off the Atlantic coast of Morocco) up to the beginning of their third winter (Fumestin and Dardignac 1962). May and June and from Nova Scotia in July to September. He estimated a 7.5% per month increase in weight of these fish during their sojourn in Nova Scotia waters. Butler (1974) supported Rivas' findings for the northern fish. He ex- amined giants taken near Prince Ed- ward Island between July and Octo- ber 1974. By calculating weekly mean weights during this period But- ler showed an approximate weight gain of 70 kg in 12 weeks. He refined these data and calculated a 30 kg per month gain between August 19 and October 4, 1974, or about 10% per month. The present authors have calcu- lated length-weight relationships for various monthly periods for bluefin tuna taken during the fishing seasons in two parts of the western North growth rate during the remainder of the year must have been much slower (about 0.8 cm per month)(Figure 8). Fumestin and Dardignac (1962) were able to collect material of ages 0-2 throughout most of the year along the Atlantic coast of Morocco. Their data for winter sizes fitted very well with corresponding data for January obtained by extrapolation from Mather and Schuck's (1960) results. It also showed that growth virtually ceased from January to March at age 0, and at the end of November, when the fish were about 63 cm long, at age 1. Data for age 2 were incom- plete, but suggested that in autumn they had practically reached their winter length of about 85 cm (Figure 9). 4. Seasonal Changes in the Length- Weight Relationship of Large Bluerm Tuna Extensive seasonal changes in the length-weight ratio of large blue- fin tuna have been found by many investigators working in different ar- eas which collectively represent a considerable part of the coastal habi- tat of the species. In the North Sea area Bahr ( 1 952), Tiews (1957) and Luhmamn ( 1 959) observed an increase in weight in relation to length during the fish- ing season of August to October. Tiews (1957) showed that tunas of 215-240 cm increased their weight during their 2-3 month stay in the North Sea by about 1 1.0 kg in 1954 and 17.4 kg in 1955. This worked out to about 34% to 54% of their yearly weight increase. Luhmamn (1959) pointed out that variations in the length-weight relationship in dif- ferent years might be directly associ- ated with variations in feeding con- ditions. Rodriguez-Roda (1964) ob- served that during the spawning sea- son, from May to August along the south Atlantic coast of Spain, mature fish lost about 1 4.73% of iheir weight between the pre- and post-spawning state. In the western Atlantic. Rivas ( 1 955) studied records of the weights of giant tuna from the Bahamas in 1.4 1.2 r 1.0 0.8 0.6 Fork Length / Moi Girth — "i^ Ftmoltt Fork Length/ Total Weight Femolet 0^ L July Aug Stpt Oct No» Figure 10. Seasonal ratiosolfork length to maximum girth and fork length to total weight by sex for all areas combined. (Caddy et al. 1976) 12 Table 6. Factors derived from the length-weight formula for Area 1 (north of latitude 35°N and west of longitude 50° W) and Area 2 (Straits of Florida and adjacent waters). Season A B Length (cm) Area 1: June -4.2571 2.7871 50-260 July -4.3893 2.8497 50-260 Aug. -4.5540 2.9290 35-270 Sept. -4.5651 2.9391 35-270 Oct. -4.7330 3.0192 35-270 July-Sept. - 4.4989 2.9044 35-270 Aug.-Oct. -4.0333 2.8606 35-270 Area 2 May-June • 4.8070 2.9044 185-260 Atlantic. Area 1 comprised waters north of latitude 35°N and west of longitude 50°W. Area 2 included the Straits of Florida and adjacent wa- ters. Factors derived from the fol- lowing length-weight formula are shown in Table 6; W = A + B Log L where: W is the live (whole) weight of the fish (kg), and L is the fork length (cm) Unpublished data collected by Schuck and Mather showed that al- though there was a marked increase in the length-weight ratio of large bluefin from June through October, the length-weight ratio of individu- als less than 100 cm long did not change noticeably from month to month during the summer. 5. Differential Growth of Males and Females Although some authors noted that male bluefin tuna attain larger sizes than females, only two studies, those of Caddy et al. ( 1 976) and But- ler et al. (1977). showing a consis- tent difference in growth rate have come to our attention. Few of the publications on the growth of the spe- cies present data in terms of sex of fish. Hamre (1960) found no signifi- cant difference in growth between males and females. Tabulations of age determina- tions by sex offish (Rodriguez-Roda 1 964) indicate that males attain larger sizes than females. The data suggest, however, that this difference is due to greater longevity for males, rather than a difference in growth rates. Rivas (1976) reported that, on the average, the males in samples of large bluefin taken off the Bahamas and in the Gulf of Mexico in various years were 4 % longer and 13 % heavier than the females. As noted in part 1 of this subsec- tion. Caddy et al. (1976) found con- sistent differences in length for age between large male and female blue- fin caught in Canadian waters in 1975 (Figure 5). These authors also found differences in the length-weight rela- tionship between males and females in the period July-November (Fig- ure 10). Males are heavier than fe- males of the same length. Since, as this figure shows, there is very little difference in maximum girth between the sexes, the males must be consid- erably heavier posteriorly than the females. 6. Ultimate Length The asymptotic or ultimate length (L_^) of Atlantic bluefin tuna has been estimated from some growth studies. Rodriguez-Roda (1964a, 1971) presented an L^ of 355.8 cm, derived from his age determinations for bluefin tuna collected off the southern Atlantic coast of Spain. Caddy et al. (1976) derived an L^of 447.88 cm from the age-size data of Mather et al. (1974), an average of several previous works, and an L of 286.64 cm for males and 277.3 15 cm for females from their own determi- nations for fish taken in Canadian waters. The longest bluefin tuna en- tered for record consideration by the International Game Fish Association as of December 1976, weighed 540 kg and was 312 cm long (FL) (the method of length measurement used was not specified) (E. K. Harry, per- sonal communication). Larger blue- fin tuna have been reported in the literature. For example. Hamre et al. (1971) reported an individual in the 310-315 cm range measured at the Istanbul (Turkey) fish market in Janu- ary 1967. Unfortunately, the weights of many large bluefin are reported without any information on their length. Sara (1969) mentioned blue- fin tuna of 625 and 685 kg taken in traps off Sardinia in 1969 and Scaccini et al. (1975) cited fish weigh- ing up to 600 kg taken in the Favignana trap off western Sicily in June 1974. 7. Discussion There has been good agreement on the size for age of Atlantic bluefin tuna for ages 1-11 (Table 4). There is less confidence in determinations for older ages. Caddy et al. (1976) extended estimated age determina- tions to 25 years, and also provided separate von Bertalanffy growth curves for males and females (Fig- ure 5). These authors note that ear- 13 lier age determinations had ascribed fish smaller than 245 cm to age group 8, whereas their own determinations indicated that 245 cm was roughly equivalent to ages 14-15 for males and age 18 for females. They also asserted that apparent underestimates of the ages offish more than 240 cm long had resulted in estimates of L^ based primarily on data from fish less than 12 years old, which were in excess of any sizes recently recorded for Atlantic bluefin tuna. If their re- sults should be validated, consider- able revisions of recent estimates of the age composition of the Atlantic bluefin tuna stocks (Sakagawa and Coan 1974) would be required. Berry et al. (1977) questioned previous age determinations for large Atlantic bluefin tuna. In the size ranges for which they provided aver- age ages (136-181 kg, 181-226 kg. 227-272 kg, 272-317 kg and 318- 362 kg), their ages were 2-3 years less than those of Mather and Schuck (1960), 3-4 years less than those of Butler ( 1 975), and 4-5 years less than those of Caddy et al. (1976). They attributed these discrepancies to the interpretation of growth marks on oto- liths and vertebrae, mainly in fish more than 10 years old. In these older fish, they often found double or mul- tiple markings which, they believed, represented subdivisions of a single year's growth. They assumed that other workers had counted each such mark as representing a year's growth, resulting in a tendency to overesti- mate ages. It is most important that the ac- tual age composition of the "relict" population (as described by Caddy et al. 1976) of giant Atlantic bluefin tuna be determined. It is also important that the lin- ear growth rate of the early stages of bluefin tuna be determined for the various spawning areas. This infor- mation is needed in terms of length, rather than weight, to permit better estimates of spawning dates and lo- calities from the collection data for these stages of the bluefin tuna. Seasonal variations in the growth rates of bluefin tuna up to 4 years old are reasonably well known (Mather and Schuck 1960, Furnestin and Dardignac 1962), but the data now available should permit extension of this knowledge to older ages. The possibility that the growth rate of the Atlantic bluefin tuna has increased as the size of the stock has decreased should also be investigated. This pos- sibility is suggested by the remarkable number of extremely large bluefin caught since 1970 (see Section IV). 14 IV. DISTRIBUTION AND FISHERIES Figure 1 1. Bluefin distribution (longline catches oft' southwestern Africa probably consisted mainly of southern bluefin, Thumms maccoyii). 90° 75" SO" 45" 30" 15* Figure 12. Bluefin regions of capture by seine, live bait, and trap 15 A. GENERAL DISTRIBUTION The bluefm tuna has been re- ported at one season or another over an extraordinarily large area of the Atlantic and the adjacent seas, and in a wide variety of water types (Fig- ures 11 and 12) In recent years it has ranged off the Atlantic coasts of Fvurope and Africa, from the North Cape, inside the Arctic Circle, to the Cape of Good Hope, and off the American coasts from Newfoundland to 40°S latitude, and also in most of the intervening oceanic areas. It also has been present in most of the adja- cent seas, the North, Mediterranean, Black and Caribbean Seas, and the Gulf of Mexico. The distribution of the bluefin tuna has vaned greatly with the sea- sons and with the size of fish (Fig- ure 13) Seasonal variations are in- fluenced by the requirements of spawning and feeding, and by water temperature. Tiews ( 1 963) concluded that distribution was limited by the \2°Q (surface temperature) isotherm. Distributional changes with size of fish are probably related to the change from a planktonic diet to one of small fishes during its first few months of life, to its first spawning at 3 to 5 years of age, or, possibly, to the fijll development of its swim bladder at ages 8 to 10 years (Sella 1929b, Sara 1973) Year-to-year changes in dis- tributional and migratory patterns are frequent Variations in environmental conditions and the availability of food are regarded as the major causes of these changes B. DEVELOPMENT OF THE MAJOR FISHERIES Large scale fisheries for bluefin tuna have existed in the Mediter- ranean and its approaches for centu- ries, but those in the remainder of the Atlantic are of more recent ori- gin Several methods and gears have been used The oldest large scale method is the tuna frap, which has been used through most of the Chris- tian era and perhaps much earlier (Pavesi 1889, Parona 1919). These were set near the coasts and the more important ones depended on the spawning migrations of the bluefin in the Medi- terranean and its approaches. An eastward or "arrival" run of prespawming fish occurs in May and June in many localities, and a westward or "return" run of spent fish occurs in July and August in some of these, and in some other locations. Most of these tuna traps were designed to catch fish coming from one direc- tion only, but a few were reversed between the sea- sons and fished both runs. Less elaborate traps have taken bluefin tuna in many localities during other seasons. Handline and harpoon fisheries have existed in many areas, in some cases for centunes, taking rela- tively small catches. Specialized gears, such as the "thonaille" (a drifting entangling net) and "seinche" 1 — 90 too LCNGTtltcml Figure 13. Lengths of bluefm tuna by region ("+" on graph means less than 0.5%). 16 (a complex multi-boat seine) of south- em France (Doumenge 1953) and the "cianciolo" (a modified seine) of Sic- ily (de Gaetani 1948), have been used locally, and generally on a small scale. Since World War II, great changes have occurred in the Allan- tic and Mediterranean bluefin fisher- ies through the mtroduction of three much more productive methods: live- bait, pelagic longline, and purse seine In addition some fisheries using the older methods were initiated or ex- panded. The traditional trolling fishery for small bluefin and albacore, Thunnus a lalunga (BonnaievTC 1788), in the Bay of Biscay was largely con- verted to the much more productive live-bait method between 1947 and 1949 (Navaz 1950a, 1950b, de la Tourrasse 1951) The French catches increased from 600 tons in 1948 to from 1 ,900 to 3,500 tons per year in 1950-1959 In 1960-1970, however, they declined to between 400 and 1 ,600 tons, exceeding 1 ,000 tons only once (Aloncle 1972). The 1970 and 1 97 1 catches were also below 1 ,000 tons (Bard et al. 1972) The live-bait method has also come into use in the fishery for small bluefin off the At- lantic coast of Morocco (Lamboeuf 1972). The highly effective purse seine method has also been widely intro- duced in the Atlantic and the Medi- terranean. The first important step was the development of a seine fish- ery for medium sized and giant blue- fin off Norway in the late 1940s fhe annual Norwegian bluefin catches increased from a few hundred tons in the middle 1940s to about 10,000 tons in the middle 1950s (Hamrc 1971). Catches declined greatly after 1952, varying from 200 to 2,500 tons per year in 1963-1971, collapsing to about 100 tons in 1972 and 1973, and rising to 800 tons in 1974 (Miyake and Tibbo 1972, Miyake and Manning 1975). A handline fishery for giant blue- fin tuna in the North Sea was origi- nated by German and Danish fisher- men in 1950. Catches peaked at 2,400 tons in 1952 and amounted to 1,800 tons in 1959 (Tiews 1975) Thus two virtually new fisheries began to take great quantities of large bluefin in the northeastern Atlantic at the same time that the only important fishery in the area for small bluefin, in the Bay of Biscay, greatly increased its catches by adopting the live bait method The German handline catches declined to 200 tons by 1962 and the fishery was abandoned after the 1963 and 1964 seasons produced only one fish each Purse seining for bluefin tuna also became widespread in the Mediterranean. Yugoslavian fisher- men evidently introduced the method in the Adriatic in 1929 (Tilic 1954) Joined later by Italian fishermen, they have seined small bluefin extensively in the Adriatic (Scaccini and Bian- calana 1959, Morovic 1961) Sein- ing of small bluefin has also been carried out by Italian fishermen in the Tyrrhenian and Ligunan Seas, but with less modem equipment Very small bluefin arc seined by sardine boats with "cianciolo" In 1972 Ital- ian fishermen began to take consid- erable quantities of giant spawning bluefin in the Tyrrhenian Sea and the Sicilian channel with large modem purse seiners (I'aini 1975) The seining of bluefin tuna off the Mediterranean coast of France was authorized in 1960 and devel- oped rapidly (di Meglio 1962). Sar- dine boats occasionally seine bluefin tuna, including extremely small ones, off the Mediterranean coasts of Mo- rocco and Spain (Rodrigucz-Roda 1964a, 1964b) Small bluefin tuna are also seined otV the Atlantic coast of Morocco (Aloncle 1964, Lamboeuf 1972) by .sardine vessels, sometimes with the assistance of chumming by live-bait boats Purse seine fishing for bluefin tuna, practiced in Cape Cod Ray by only one or two very small vessels in 1958-1961 (Squire" 1959), expanded to an oceanic fishery ranging from Cape Hatteras, North Carolina, to Cape Cod, Massachusetts, in 1962 (Wilson 1965) The catch of small bluefin rose to 5,600 tons in 1964, along with an almost equal quantity of skipjack tuna, Katsuwnnus pelumis (Linnaeus 1758). The largest fleet in the history of the fishery, 2 1 vessels, including some of the world's larg- est, was responsible for this catch. Subsequent annual catches have var- ied considerably between a low of 600 tons in 1966 and a high of 3,600 tons m 1 97 1 The Japanese longline fishery en- tered the Atlantic in 1956, and subse- quent expansion was rapid (Shiohama et al. 1965). By 1962, most of the ocean between latitudes 25°N and 25°S was being fished by Japanese longliners In 1969, the fishery had expanded so that most of the waters between 40°N and 40°S were being fished (Wise and Davis 1973). The catches of bluefin tuna were small (less than 7,000 fish per year) in 1956- 1961, but increased to 53,000 to 67,000 fish per year in 1962-1965. They then declined to less than 1 ,000 per year in 1969 and 1970. The 1971 catch was 8,000 fish, however, and tlie catch had risen to about 46,000 fish by 1974 (Fisheries Agency of Japan 1 976). This was due mainly to the entry of the fishery into areas which had formerly been unexploitcd, or exploited only by inshore gears — the oceanic eastern Atlantic, the Bay of Biscay, the Ibero-Moroccan Gulf and the Mediterranean Japanese longliners entered the bluefin tuna fishery in the Bay of Biscay in 1974, reportedly affecting the operations of the local Spamsh fleet (Cort and Cendrero 1975). Their catches in the area in that year to- talled about 11,215 fish (Fisheries Agency of Japan 1976). Japanese longliners entered the Mediterranean in 1972 Their total catch that year included 1 12 tons of Atlantic bluefin tuna. This increased to 246 tons m 1973 and 2,1 95 tons in 1974 (Miyake and Manning 1975). Much of this catch was taken in the vicinity of Sicily in June and July, during the spawning sea.son (Shingu et al. 1975, Shingu and Hisada 1976, Fisheries Agency of Japan 1975, 1 976) 1 he concentration of longlines around Sicily reportedly made it al- 17 most impossible for the Italian sein- ers to set their nets (Paini 1975). In 1975 the Japanese government pro- hibited their longliners from fishing in the Mediterranean as part of its compliance with the ICCAT regula- tions (Kume 1976). Numerous longline vessels of South Korea, Nationalist China, Cuba, Venezuela, and the Soviet Union ha\e also caught bluefin tuna, along with other species, in the At- lantic in recent years (Miyake and Tibbo 1972). Offshore big game fishing for bluefin tuna and other large oceanic fishes (chiefly billfishes) became popular in parts of the northwestern Atlantic during the 1 930s (Farrington 1939) and expanded to a highly de- veloped pursuit after World War II (Farrington 1949) Centers of in- tensive sport fishing for bluefin tuna have been off the northwestern Ba- hamas in spring, and along the United States coast from Virginia to Maine and in Canadian waters off south- western Nova Scotia, eastern New- foundland, and in the Gulf of St Lawrence in summer and early fall Thousands of private and charter boats have participated in this fish- ery, with as many as a thousand an- glers in two hundred boats entering a single tournament. Sport fishing for bluefin tuna has also been popular in eastern Atlantic and Mediterranean waters. Centers have been the Oresund off Denmark, the North Sea off England, the Canary Islands, the Bav of Biscay oft' northern Spain, the Italian Riviera, and the Mediterra- nean coasts of France and Spain Re- cent low availability of fish has re- duced interest in sport fishing lu sev- eral of these areas The northeastern Atlantic fish- eries for large and medium sized blue- fin tuna began a decline in 1963 which reduced several of ihcm to ex- tinction bv 1973. Bluefin tuna pro- duction in the Japanese oceanic longline fisher> peaked rapidly in 1 962- 1 965 but declined to a low level by 1967 The reported total catches of the Mediterranean fisheries, on the other hand, appeared to be maintaining themselves satisfactorily as iccenlly as 1971. Apparently, however, cen- tral Mediterranean (Italian. Tunisian, and Libyan) trap fisheries, have re- cently undergone a collapse similar to that experienced by those in the eastern Atlantic. Catches were fairly good through 1969, but in 1972 and 1973 traps which had fonnerly pro- duced thousands of fish per year yielded only a very few hundred Only a handful of the once numerous Italian traps survive. This decline has been offset recently by the entry of the Japanese longline fisher^' into the area and the development of the Ital- ian purse seine fishery for giant blue- fin in the central Mediterranean. It should be noted that both of these fisheries also use spawning concen- trations Another notable trend in nearly all of the Atlantic and Mediterranean fisheries for large bluefin for vhich data are available has been the \irtu- allv complete absence of medium sized (32-122 kg, ages approximately 5-8) and a marked scarcity of small giant (123-200 kg, ages approxi- mately 9-11) bluefin from the catches This trend has been reversed in the central Mediterranean fisheries in 1975 (Miyakc 1976) In the following section, we will discuss the distribution of Atlanfic bluefin tuna and the fisheries for it in detail. C. DISTRIBUTION AND FISHERIES BY AREAS 1. Introduction In this section we will discuss the following information, bv area; 1 ) Times and locations of occur- rences of bluefin tuna, ar.d Ihcir \ ana- tions with size of I'ish 2) Fisheries and fishing meth- ods. 3) fhe volume and si/e compo- sition of the catches, and their trends The size composition by years of samples of the catches of most of the ma]or fi.shcnes is illustrated b\ histograms. The tables from which most of these histograms were plot- ted have been reproduced, with the sources of the data, m ICCAT Data Record, Vol. 3, Madrid. 1974 2. Western Atlantic ■fhc maior fisheries for bluefin tuna on the western Allanlic conti- nental shelf are concentrated between Cape Hatteras, North Carolina, and Newfoundland, Canada (Figure 14). The small fish have been predomi- nant in the area between Cape Hatteras and Cape Cod, Massachu- setts, but medium sized and giant fish have also occurred there in num- bers on occasion. Giant fish have been predominant m the region between Cape Cod and Newfoundland in re- cent years Until 1966, medium sized bluefin were often abundant between Cape Cod and Nova Scotia. In some years during the 1950s, small bluefin were the most numerous size group in northern Massachusetts waters, but wc have nc\'er heard of their occur- rence in significant numbers off Canada The gears used for bluefin tuna in this area include harpoons, hand lines, traps, rods and reels, and purse seines Some giant bluefin have also been taken by sport gear oft' the north- western Bahamas and by small-scale commercial fisheries off some Cu- ban ports Captures of bluefin tuna elsewhere in western Atlantic coastal waters have been insignificant. a. Newfoundland to Cape Cod i. Newfoundland There has been little commer- cial fishing for bluefin tuna off the Canadian island of Newfoundland (Figure 15), but good sport fishing for the species was enjoyed there from 1961 through 1972. This fishery was initiated at Con- ception Bay in 1956. and was .veil developed there by 1961. Another productive ground, Notre Dame Ba\ , was opened in 1967, with most of the boats ba.sed at Lcwisporte 'fhc fish- ing usuallN extended from mid-July to mid-October from 1961 through 1972. catches of up to 635 fish were taken in Newfoundland waters each season Many of these fish were tagged and relea.sed. The catch has declined drastically since 1972, with less than 100 fish being taken each season Nearly all of the catch con- sisted of giants, and their modal 18 50° - 40* - 30" 20° 10° 0° <:^^., ' - "~-. 5 CAPE COO . Xj'/t'^T^*© ""-, ' 6 GEORGE'S BANK j|te»"-. '- »*•«> / . I » ? 7 MIDDLE ATLANTIC BIGHT ^^^/"^^^^ ^■''—-'^""^ ' _ NEW yobkW^ .6' - • . -3© C*'E NATTtRAS ...■:::,-.:■., -rp "x^"" ' '^^^^^^^'''" '\\V\' ■j;' GULF OF \ ^^^rA. f -"'"-. ;^^ - ■ l, —- ' 1 J^^i;>~;». ,■•••. \. ■ -r^. •*>.^''>%V/ ,r-. - ""^aiK^'Li^ii^- ■-;■•■" '"^"-^Es' ';-i >*4^, ■ ^ /' CARIBBEAN SEA | '^ ^ - ^Xj * '' -^"i^' ^' ^ t \ J!f^'''^if^'^k*A- '*■" "^ "■ ^^^?^;»^^^v;ir ^' n'*V \ ^^Y' ■-.^'S,,,^^-,^ J- ^^"'*^'- " ^- ^15 ' 100° 90° 80° 70° 60° 50° Figure 14 Geographical references in the northwestern Atlantic 40° length increased from about 220 cm in 1961 to about 240 cm in 1971 (Figure 16). Their mean weight rose hkewise, from 220 kg in 1 961 to 3 14 kg in 1975 (Caddy and Butler 1976) ii. Gulf of St Lawrence In recent years, more extremely large bluefin tuna have been taken on rod and reel in the Gulf of St Lawrence (Figure 17) than in any other area. No less than 36 out of 487 tuna caught in the Gulf in 1975 ex- ceeded 1 ,000 lbs (454 kg) in weight (D A. MacLean, personal communi- cation). The sport fishery was initiated at Prince Edward Island in 1966 The season originally extended from late July into October, but the high prices offered for large bluefin tuna caught in the late season resulted in its ex- tension into November, and also in an increase in the number of boats fishing from 30 in 1972 to about 72 in 1973 (Anonymous 1974). Almost all of the fish taken have been giants (Figure 18) The average weight of the fish caught has remained consis- tently over 300 kg, and rose sharply to 382 kg in 1 975 (Caddy and Butler 1976). With increasing fishing ef- fort, the total catches rose to a peak of 1,048 fish in 1974, but declmed sharply to 343 fish in 1975 (D A MacLean, personal communication) In 1972, the Canadian Minister of Fisheries restricted bluefin tuna fish- ing in the Gulf of St Lawrence to rod and reel, using lines of not over 130 lb (59 kg) test (Anonymous 1974) In 1975 the Minister restricted the fishing to 10 week seasons and the catch to two fish per boat per dav, using rod and reel only In addition, licenses were limited to boats which had fished in 1974. These restric- tions, which were part of Canada's compliance with the ICCAT regula- tions, combined with poor abundance of fish to reduce the catch in 1975 (Caddy and Burnett 1975) The success of the Prmce Ed- ward Island fishery and exploratory efforts by visiting United States an- glers resulted in the extension of the sport fishery in 1973 to Chaleur and Gaspe Bays, New Brunswick. This fishery increased rapidly and took 93 fish with an average weight of 801 lbs (363 kg) in 1974 and 148 fish with an average weight of 858 lbs (389 kg) in 1975 (D A. MacLean, personal communication) Potential world record rod and reel bluefin weighing 1 , 1 90 lbs (540 kg) and 1 ,200 lbs (544 kg) were taken in Chaleur Bay in 1976 (E K. Harry, personal communication) The continuing increase in the sizes of the bluefin taken in the Gulf of St Lawrence, and the lack of re- cruitment to the fishen', are vividly illustrated by the weight composi- tion of the 1974-1976 catches (Fig- ure 17) iii. Nova Scotia Bluefin tuna have long been taken by traps and harpoons in Nova Scotian waters, but the only histori- cal size data are from the sport fish- ery. The sport fishery was first de- veloped in 1935 at Wcdgeport south of Yarmouth) (Figure 19) where it was very successful for many years. In the mid 1950s, however, the fish- ing there deteriorated, and Nova Scotian fishing has been centered at Cape St Mary (north of Yarmouth) since then Our data extend from 1 946 tlirough 1955, and also include 1959. The size composition of the catches \aned considerably over these years (Figure 20) Giants dominated the catch in 1 946, but medium sized fish became more important in succeed- ing years, dominating the catch in 1949 and pro\iding over 40 percent of those in 1948 and 1950 The year class of 1942 was dominant in 1948, and that of 1943 was dominant in 1949-1950, and important in 1951 (Mather and Schuck 1960). Giants again dominated through 1955, ex- cepting an important showing of me- diums (year class of 1949) in 1954. In 1959 mediums were again domi- nant (>ear classes of 1952 and 1953). The few catches in recent years have 19 54° 53° 52' ^o - 47° 46° 58° 56° 54° 52 A ' 1 ' ^ 1 1 1 CITIES ^ ®^ *>® NEWFOUNDLAND ,6*' - ,2 _ // MiOUELON IS E Ojy:>' CAPf BRE TON I 5LAN0 /' _l L- Figure 15. Geographical references off Newfoundland and Labrador been of very large giants. Their aver- age weight has also increased in re- cent years, reaching 761 lbs (.'^45 kg) in 1974 and 824 lbs (394 kg) in 1975 (D A. MacLean, personal communi- cation) The fishmg off southwestern Nova Scotia usually extended from early July to late September or Octo- ber. The giants have been most abun- dant in August and September, but the medium fish, which arrived and departed later than the giants, have been most abundant in September and October. Nova Scotia has been the locale of the prestigious International Tuna Cup matches since their incep- tion in the 1937 season. As docu- mented by Farrington (1974), these competitions were held at Wedgeport through 1957, except for one at l.iveipool and an interruption caused by World War II Subsequent matches were held at Cape St Mary in 1958 and from 1965 through 1975 niuefin tuna have been an im- portant incidental catch in the mack- erel trap fishery at St Margaret's Bay (near Halifax) Nova Scotia (Figure 19) Young (1974) briefiy reviewed the history of the fishery, and Butler (1977) has provided details on its recent trends and developments, and Its status in 1976 Additional data are available in Butler (1975), Caddy and Butler (1976) and statistics of Envi- ronment Canada (D A. MacLean, personal communication) Young (1974) stated that land- ing statistics for the fishery have been recorded since 1918, but the fishery predated the establishment of statis- tical systems The recorded annual catches of large fish have been in the order of 400, but have ranged as high as 1,500 There was also a second later run of much smaller fish, m the 20-70 kg range. None of the small fish have been seen in recent years, and the average weight of larger fish increased con- siderably, reaching 295 kg in 1974 ■fhe weight composifion of the 1973- 1975 catches (Figure 21) shows a considerable increase in sizes ol fish taken each year, and only fragmen- tary recruitment of younger fish to the fishery The estimated numbers of fish caught in the period 1964-1975 var- ied between 104 (in 1972) and 865 (in 1974) (Butler 1977) In June 1974 an additional 150 tuna caught in the traps were tagged and released Ton- nages in the years 1971-1975 ranged from 23 6 mt m 1972 to 256 mt in 1974 (Caddy and Butler 1976, Butler 1977, D A MacLean, personal com- munication). The average weight of the fish increased from about 227 kg in 1971-1973 to 295 kg in 1974 and 319 kg m 1975. Butler (1977) described a most interesting development in this fish- ery — the holding of giant bluefin in impoundments, and fattening them, to secure optimal prices. Japanese interests offer verv' high prices for fresh giant tuna in the fat condition which thev attain in late summer and fall through heavy feeding The prices offered for the typically lean fish taken in the early season are much lower Butler stated that the average price per pound offered to Canadian fishermen for giant bluefin in 1976 varied from $0 20 m July-August to $ 1 40 in October. The St. Margaret's Bay traps normally take most of their catch of giant tuna in the early sea- son, when the price is low. In an 20 30 20 <0 30 20 K 10 1960 -ff- SO 20 10 30 1d6t 1962 ^Z- 20 10 h 1965 i-ff- 30 20 10 1964 30 r 20 10 \- -a- 1965 h^T- 50 . 1966 -a- 1967 'r?J- Jk - 1966 i-^?- 4 1969 -l^ t970 jL 1971 too 150 200 250 300 50 100 LENGTHkm} Jl 1 jL I I I i^ 150 200 250 JOO LENGTH (cm) Figure 16. Lengths of bluefin tuna captured ofl" Newfoundland (1960-1971) ("+" on graph means less than 0.5%). 21 80 60 40 20 1976 - ^^- Caiaquel I I - Prince Edwraid Island il 40- 20- 80 60- 40 20 100 -f=JS^ p ThvfT - -"■H 1975 ^ f -^ ral ^3 i - 1 R'pq . 1974 iU ni 200 ^1 ? '^--^i 2 1^ 100 400 W«ighl ( 10 kg groupings) 500 Figure 17. Weights of giant bluefin tuna caught by rod and reel in the Gulf of SL Lawrence (1974-1976). methods in the New England area (Figure 22) (Bigelow and Schroeder 1953, Wilson 1965). The traps, which were formerly numerous in Cape Cod Bay, took considerable quantities of these tuna m years of abundance, but the fishery there was termmated in 1975. Operators of various types of commercial fishing craft, such as lob- ster boats and small trawlers, occa- sionally harpooned bluefin or caught them with handlines. This activity was usually incidental to their nor- mal pursuits or undertaken tempo- rarily during periods when tuna were readily available and commanded a reasonable price. Sport fishing for bluefin tuna developed in the area in the 1930s (Farrington 1939) and increased in popularity after World War II (Farrington 1949). Pre-war activity centered in Ipswich Bay, Massachu- setts, and Casco Bay, Maine. After the war the popularity of sport tuna fishing spread into Massachusetts and Cape Cod Bays as well. Some of the most successful giant tuna tourna- ments ever held have taken place in, effort to capitalize on the high late season prices, two impoundment nets were constructed in the Bay in 1975. Each impoundment measured about 100 m X 50 m, and was in water about 20 m deep. Fifty bluefin were trans- ferred from neighboring traps to the pounds. They were fed on trash fish and held for a period of two or three months. Then they were killed in small batches, to avoid depressing the price, and sold on the Japanese mar- ket. The experiment was so success- ful that nine impoundments were con- structed m 1976, and about 300 blue- fin were held and fattened. Butler (1977) showed the loca- tions of the traps and impoundments in St. Margaret's Bay and discussed past and proposed research activities in this unique situation iv. Cape Cod to Maine Bluefin tuna have occurred in many sizes and been fished by many 30 r 20 10 - 30 1- 20 30 20 - 10 - 150 200 LENGTH icm) 250 300 Figure 18 Lengths for bluefin tuna captured off Prince Edward Island (1969- 1971). 22 Figure 19. Geographical references of Nova Scotia. these waters in the years 1972-1976 Large bluefin were remarkably avail- able in those years, and their average size was increasing steadily. The purse seine fishery for blue- fin t\ma off the east coast of the United States onginated in Cape Cod Bay and its vicinity (Wilson 1965) After initial expenments in 1951 and 1954, a continuing purse seine fishery be- gan in 1958 with a single small ves- sel operating out of Provincetown, Massachusetts. Very high catch rates were obtained from the schools of medium sized bluefin which were then abundant in the Bay (Sakagawa 1975) In 1962 the carrying capacity of the fleet was greatly increased by the purchase of additional vessels by local fishermen and visits by larger ones from other areas, and the fish- ing area was expanded southwest- ward to include the New York Bight and more southerly areas The large fleets which entered the fishery in 1963 and 1964 made considerable portions of their catches of small and medium bluefin in Cape Cod Bay and vicinity, especially toward the end of the season. Very few small or medium fish have been found in the Gulf of Maine since 1964, For sev- eral years thereafter, seining in the area was limited to occasional trips by one or two vessels after the sea- son for small bluefin southwest of Cape Cod had terminated This situ- ation was altered by the high prices oftered fi-om 1972 on by Japanese interests for the fat bluefin which were caught late in the season A local seiner, who specialized m catch- ing giant fish, had fished in the Bay for the Japanese market toward the end of each season since 1972 In 1976 a second vessel joined the fish- ery. The prices offered for large blue- fin tuna also caused a dramatic in- crease in fishing effort by the har- poon, handline and sport fisheries The "sport" fishery differed only in the gear used, since virtually all of the anglers were fishing for the mar- ket Boats of all types and sizes, from small outboard motor boats to large trawlers and party fishing boats, par- ticipated Mather and Mason attempted to gather catch records during the be- ginning of this period of increased effort and estimated catches of 3,000 and 3,500 fish in 1972 and 1973, respectively In 1974, the Commonwealth of Massachusetts regulated the fishery for large bluefin in its waters and collected data on the landings. The National Marine Fisheries Service also initiated a data collection pro- gram for the fishery. The resulting data are shown in Table 7 (F. H. Berry, personal communication). The data for the rod and reel and harpoon and handline catches were estimated Since 1975 quotas have been im- posed on these fisheries to control the number of large fish taken each year and to prevent the capture of the surviving small fish (National Ma- rine Fishenes Service 1975, 1976), Under the 1975 quotas, the recorded catch by hand gears (harpoon, handline and rod and reel) was 2,277 fish weighing 693.5 tons, with an average weight of 305 kg. The re- corded seine catch was 1,017 fish weighing 267.2 tons with an average weight of 263 kg (Aloncle et al 1976, ICCAT 1976) The time of the best fishing in Cape Cod Bay has vaned consider- ably — partly according to the sizes of the fish which entered the area. The giant fish usually amved early, in June or early July, and stayed into September or October. Peak fishing was usually in August or September. Medium sized fish have usually arrived later, and also stayed in the Bay later than the others, being abun- dant in August-October. Exceptions occurred when the smaller (age 5) individuals of this group accompa- nied early summer runs of small (age 2-4) bluefin Small bluefin were un- predictable Some strong runs peaked in early summer, as in 1 95 1 and 1953, but others ha\'e occurred in late sum- mer, as in 1950 As noted above, fishing tech- niques in the area have changed over the years Also, catch records have not been kept in a consistent manner We do have sporadic records of the sizes of fish landed for Ipswich Bay and the Maine coast for 1947-1951 (Figure 23) and more comprehen- sive ones for Cape Cod Bay and vi- cinity (including Ipswich Bay Irom 23 >0 20 • O O 30 20 10 1946 1947 L-^?- > 20 «948 30 20 U l^'*^ 10 - o so 20 10 30 20 10 1992 . 1963 1950 1961 -J^ SO I lOO ■^Jf- 1954 . 1955 1959 -UU nt 1 r 50 100 150 200 L£NaTH(cm) — r soo 150 200 LeNGTHUm} 250 Figure 20. Lengths for bluefin tuna captured off Nova Scotia (1946-1959) ("+" on graph means less than 0.5%) 3C0 WEIGHT Ikg) Figure 21. Weights of bluefin tuna captured from St Margaret's Bay ("+" on graph means less than 0,5%), 1960 on) for 1947-1975 (Figure 24). As the latter figure shows, the size composition of the samples has var- ied drastically over the years. In 1947, giants were dominant, with a fair pro- portion of mediums m Cape Cod Bay. The mediums, mainly the year class of 1943, which was also outstanding in the Nova Scotia catches of 1949- 1950 (Mather and Schuck 1960), be- came dominant in Cape Cod Bay in 1948 and 1949. They would have been dominant in 1950 also but for an apparently unprecedented inlTux of small bluefin. 'I'his 1948 year class remained dominant through 1957. The mediums became dominant again in 1958-1960, mainly because of the progression of the veiy strong year class of 1952 through this size cat- egory. This year class was also promi- nent in the 1959 Nova Scotia catches. In 1961, small, medium, and giant bluefin were all represented, but an- other strong year class (1957) had become dominant fhe progression of this class made the medium group 24 72° 70° 68* 66° 45° 44« 43° 42° 1 1 ^ CITIES ® GEOGRAPHIC REFERENCES 1 CAPE COD BAY 2 MASSACHUSETTS BAY 3 IPSWICH eAY 4 BOONE ISLAND 5 OGUNQUIT 6 eOOTHBAY HARBOR 7 PEMAOUIO POINT B MUSCONGUS BAY o MOUNT DESERT I BAKERS L 10 GRAND MANAN I 11 PASSAMAOUOODY BAY 12 ST JOHN 13 DIGBY PO«TL» ''' y^p< Figure 22. Geographic references in the Gulf of Maine region. preponderant in 1962-1963. Giant tuna have dominated the catches in all the subsequent years (1 964- 1 975), with only negligible numbers of small and medium sized bluefin being re- corded between 1966 and 1970 As in Newfoundland, the modal lengths of the giant bluefin taken in this area in the late 1960s reached high val- ues, over 220 cm, and the size of the smallest fish taken increased signifi- cantly. The latter trend changed in 1970-1971, indicating some recruit- ment of younger fish. In subsequent years, however, the catches have con- sisted almost entirely of very large fish, and the mean size has been steadily increasing. Catches of blue- fin less than 8 years old have been insignificant. The available data for 1973-1975 show a continuing increase in the size of the fish taken, especially by the hook and line and harpoon meth- ods The average weights of fish caught by nets (traps and purse seines) are somewhat lower, possibly because of the tendency of smaller fish to school together in greater numbers than the larger ones There has been a fairly close re- lationship between the occurrences of medium sized bluefin in Cape Cod Bay and off Nova Scotia, but the important runs of small fish which have occurred in the former area have had no counterpart m the laUer The runs of small bluefin in Capo Cod Bay have comprised ages 2-4, and we are aware of only one indi- vidual of age 1 having been reported from this area The age composition of these size groups has often over- lapped, with age 4 fish associating with runs of medium fish, and age 5 fish with runs of small bluefin b. Cape Hatteras - Cape Cod (Coastal Waters) The "Middle Atlantic Bight" (Figure 25) is the usual summer habi- tat of the small (age 1-4) bluefin of the western North Atlantic, although they have occasionally ranged into the Gulf of Maine (chiefiy Cape Cod Bay) as noted above Many traps in this area took bluefin tuna occasion- ally, but very few of these are now in operation The bluefin tuna has been one of the most abundant and popu- lar big game species supporting the offshore sport fishery of this area, particularly in its northeastern part. Since 1 961 it has also been the object of a seasonal purse seine fishery which also takes skipjack tuna in some years. Bluefin usually arrived in the area in late June or early July and departed in the early fall, but they sometimes departed as late as November. The size composition data available to us for this area (Figure 24) are mainly from the sport fishery for years before 1961, and from the purse .seine fishery for later years. As indicated by these histograms, the sport catches ha\e consisted al- most entirely of small bluefin, mostly of ages 1-3 Sport fishing for these bluefin has been very successftil in some years off New York Harbor (Wcstman and Neville 1942, Moss 1967) and elsewhere between New Jersey and Cape Cod (Farrington 1939, 1949). The best season is un- predictable. Sometimes the early sea- son produces the best fishing, whereas in other yeais there is excellent fish- ing in late summer and early fall, particularly in the grounds off New York Harbor. The purse seine fishery began to harvest this stock in 1962 In the first three seasons (1962-1964) consider- able quantities of medium sized blue- fin were taken, but in subsequent years, fish of this group have been rare in this area, as well as in the more northerly waters which they formerly frequented (see above). Trends in the catches are shown in Figure 27 Wilson (1965; and Sakagawa (1976) give details of this fishery and its catches The size composition data (Fig- ure 26) suggest some variation in the sizes taken by the various gears In the years when most of the data were from rod and reel catches. 1 94 1 - 1 960, age 1 fish were often important. In the years from 1962 on, when the seine fishery was the chief harvester of young bluefin in this area, age 1 fish were important only in some years, notably 1 966 In the first three years (1962-1964) of intensive seine fishmg, fish of ages 4 and 5 were 25 30 r 20 10 - 1947 •— ^-^ 30 r- 20 10 1948 Uj 30 '^ ^ 20 l^^ 10 « o 30 20 10 O 30 20 10 1949 1950 1951 50 „i ±±, 100 150 200 LENGTH (cm) 250 300 Figure 23 Lengths of bluefin tuna captured in Maine and Ipswich Bay, Massachusetts ("+" on graph means less than 0.5%). 26 important Many additional fish of ages 5 and over were taken in thiis period but were discarded because they were not acceptable to most of the canneries (several eyewitness re- ports, personal communications). Also many fish were lost when nets burst, or were discarded when the catch exceeded the remaining carr)'- ing capacity of the vessels (the data sources are the same as above, in addition to incidents witnessed by some of the authors and their col- leagues) After 1966, fish of age 2 usually dominated the catch. Since 1974 the New England based seiners have generally avoided age 1 fish as a voluntary conservation measure Since midsummer 1975, their cap- ture has been prohibited in conform- ance with the ICCAT regulations It appears, however, that in most years full recruitment to the fishery does not occur until age 2. The size composition data give indications of strong-year classes The year class of 1938 was dominant in the New York Bight in 1941 in 1 95 1 the year class of 1 950 showed up very strongly in the catches, but no sampling was earned out It was very prominent, however, in the 1952 sample. The class of 1958 was con- spicuous in the 1960 and 1962 samples The class of 1 965 was domi- nant in the 1966 seine fishery, and showed up strongly in 1 967 and 1 968 Data for 1974, 1975, and 1976 indi- cate a very strong year class of 1973. Sport fishing for giant bluefin off New York Harbor was fruitftil in the 1930s (Farrington 1939), but de- clined after World War II Late in the 1940s, a new giant tuna ground was discovered near the western coast of Rhode Island (Farrington 1949) This was productive for some years, but after 1960 successfiil seasons became infrequent More recently, giant tuna have been caught farther off the Rhode Island and eastern Long Is- land shores, in areas where the tunas follow the nets of trawlers as they are being hauled to the surface This fish- ing has usually been most successful in late summer and early fall After being absent for many years, some giant bluefin have been taken off northern New Jersey or during the later part of some recent seasons. Table 7. Landings data from the Commonwealth of Massachusetts in 1974. Number Metric Average Gear offish Tons Weight (kg) Trap 37 9.1 245.6 Seine 167 48.1 287.5 Rod and reel 200 63.5 317.5 Harpoon and hand line 900 349.2 317.5 These sport catches of giant tuna were usually unimportant in numbers, but did document their seasonal occur- rence in the area. At the other end of the size range, runs of very small (age 0) bluefin (1- 2 kg) occasionally occur in this area from late July into October. These are most common in late summed and early fall, and from northern New Jersey southward to Cape Hatteras They are of no significance to the fishery, but the biological informa- tion obtained from them has been a ver\' important step in understanding the life history of the species c. Atlantic Ocean outside 200 meter contour N of 35°N andWof40°W During the cold season, the blue- fin tuna disappear from the summer habitats described above Since 1956, however, exploratory and commer- cial longline activities have revealed much about their oceanic distribu- tion in this period (Wathne 1959, Wilson and Bartlett 1967, Wise and Davis 1 973) The only substantial size composition data for oceanic bluefin tuna catches are from the area north of 35°N and west of 40°W (Figure 28) The most important exploratory catches were near the northern edge of the Gull Stream in spring, and in the canyons along the edge of the continental shelf in fall (Mather and Bartlett 1962). I'he majority of the fish taken were in the medium (120- 1 85 cm) size range As in the inshore fisheries, however, the average size of the fish taken, as well as the size of the smallest fish taken, increased markedly during the 1960s. Japanese longline catches in this area have indicated a predominance of the larger medium sized fish (about 100 kg) during the period from April through October (Shingu et al. 1975). Shingu and Hisada (1976), however, reported that medium-sized bluefin tuna (ages 5-8) were scarce through- out the Atlantic Ocean. They also noted that, during the winter months, December-February, considerable proportions of small fish (ages 2-5) have appeared in the catches of re- cent years in the waters off New En- gland d. Bahamas and Southeastern Florida I'he only bluefin fishery in the western Atlantic south of 35°N for which there is significant size com- position data IS the sport fishery off the northwestern Bahamas (Cat Cay and Bimini) (Figure 29). All the fish taken are giants, over 185 cm and 122 kg (Figure 30), and are caught in May and June Usually the fish are caught from schools which are trav- elling northward along the edge of the Great Bahama Bank, but on rare occasions they are taken from schools which are "smashing" (feeding on the surface, often jumping clear of the water) farther offshore. This fish- ery has varied somewhat in recent years, with a general tendency to de- cline Many of the fish have been released after being brought to the boat We have some weight records for more recent years, however, which, although few in number, con- stitute representative samples These samples indicate that the modal sizes and minimum sizes of the fish taken in this fishery have increased steadily in recent years Rivas (1976) showed that the mean length of the males which he had measured in the years 1972-1973 at the Bahamas was 25 cm longer than the mean of those which he had measured in the years 1952-1 955 For the same periods, the 27 200 i £^GTH fen* J laO ZOO 30 za to i so 10 O ' — I < o ■^ 10 50 20 »o 30 eo 1 o so f- zo «o O ' iJ-^-f- x. iOO * &0 200 2&0 SOD SO :er.o SOO i erAfGTf^ fCrrr? Figure 24. Lengths of bluefin tuna captured in Cape Cod, Massachusetts bays and vicinity (1947-1975) ("+" on graph means less than 0.5%). 28 Figure 25. Cicographic references for the Cape Hallaas to Ca|K; Cod areas. mean length of the females which he had measured mcreased by 20 cm Similar fishmg occurs, with fewer boats participatmg, along the west- em edge of the Little Bahama Bank (West End to Malanilla Shoal). Giant bluefm have also been taken off the eastern Bahamas, from Cat Island to Walker Cay (Anony- mous 1962), in the same season. A few have been taken off southeastern Florida on rare occasions during the winter and spnng Small and medium sized bluefin are very rare or nonexistent in the vicinity of the Bahamas, but runs of very small (less tiian 2 kg) fish some- times occur off southeastern I'lorida in July-September, and individuals are sometimes taken through the fall and into early winter (Rixas 1954, Mather and Schuck 1960, Mather 1963) c. Gulf of Mexico and Caribbean Sea Bluefin tuna are caught b) longline, and occasionally bv sport fishing, in the Caribbean Sea and the Gulf of Mexico (Figure 31) Catch records, aside from those of the Japa- nese longline fisheiy, and size data are sparse, but sulTicient to give a good idea of their distributional pat- tern Most of tlie catches ha\e been m the first two quarters of the year United States exploratory catches (Wathne 1959, Anonymous 1962) and commercial catches (H R Hullis. .Ir , personal communication) ha\e been mamly in the Gulf of Mexico and the northwestern Caribbean The Japanese effort, which has been much more extensive, has shown that the species occurs occasionally through- out most of the Canbbean (Fisheries Agency of Japan 1967a) A small scale Cuban longline fishery for giant bluefm has operated in the spring, over the Bartlett Deep between Cuba and Jamaica. This fish- ery was initiated in 1 969 and the larg- est catch of about 360 tons (includ- ing incidental catch of other species) was taken in 1 970 The landmgs were smaller, but good, in the next three years, but the 1974 catch was ex- tremely poor (Ubeda 1974). Cuban handliners occasionally catch giant bluefin off Havana in the spring (Rivas 1954), All of the longline catches in the Gulf of Mexico and the Caribbean Sea for which we have size data were giants, over 185 cm long. Similar size fish have been encountered oc- casionally by sport fishermen be- tween Cozumel Island and the Yucatan Peninsula and off the Mis- sissippi delta in late spring on the surface, sometimes in large schools (Nakamura and Rivas 1972). We had \'irtually no records of occurrences of medium sized or small bluefin tuna in these waters until Shingu et al. (1975) reported catches of bluefin tuna weighing less than 10 kg b\' Japanese longline vessels in the Gulf of Mexico in June and July 1973. We have had several reliable reports and records, however, show- ing that very .small (age 0, 2 kg or less) individuals have occurred in the Gulf of Mexico from July mto No- vember 3. Atlantic Oceanic Waters Hx tensive data on the seasonal distribution of bluefin tuna in the oce- anic waters of ihc Atlantic have be- come available from the records of the Japanese longline fishery (Shiohama et al. 1965, Fisheries Agencv of Japan 1965, 1966, 1967a, 1967b. 1968, 1969, 1970, 1971, 1972, 1973, 1974, 1975, 1976, Wise and Lc Guen 1969; Wi.se and Davis 1973) as well as from exploratory fishing (Wathne 1959, Anonymous 1962, Wilson and Bartlett 1967). The gen- 29 LENCTH Figure 26. Lengths of bluefin tuna captured from Cape Hatteras to Cape Cod ("+" on graph means less than 0.5%). 30 4- 3- 1- 0- □ NUMBER OF FlSH/lO* ■ metric tons/io' 1958 Figure 27. Annual catches of small and some medium size bluefin tuna from the northwest Atlantic purse seine fisheries, 1958-1973, in tens of thousands offish and thousands of tons. era! seasonal distribution pattern, based on Japanese catch rates for 1956-1969, is illustrated by Figure 32 (Wise and Davis 1973). This fig- ure, however, includes southern blue- fin, Thunnus maccoyii, as well as the Atlantic bluefin, 7" thynnus thytmus. The records south of latitude 2()°S, and perhaps also the one in the first quarter centered at 12°S, 2°E, are probably of southern bluefin. Catch rates for individual years and areas (Table 8, Figure 35) (J P Wise, personal communication. Wise and IDavis 1 973) suggest that the abun- dance has varied considerably over the years. The catches declined pre- Figure 28. Lengths of bluefin tuna captured in the Atlantic tail.sidc the 2(K) meter amtour north of 35°N and west of 40° W ("+" on graph means less than 5%) 31 Figure 29. Geographic references for the Bahamas and southeastern Florida. cipitously from 1965 through 1970 but have risen rapidly since then (Fig- ure 33). The following description is based on the 1 956-1 968 catches (Fig- ure 32): In the first quarter of the year, the highest catch rates occurred around the easternmost part of Bra- zil, and the greater Antilles In the second quarter, the area of relatively high catch rates off Brazil extended farther to the northwest, and catch rates were high in the northern Gulf of Mexico, and off the Bahamas and the eastern United States north- ward to latitude 40°N and eastward to longitude 65°W. The concentra- tions off easternmost Brazil and off the Bahamas appeared to be con- nected (Figure 32, Appendix Figure 62 in Fisheries Agency of Japan 30 20 10 30 20 10 t«l 30 « 30 zo 10 50 20 10 (939 -Z^- 1947 H948 ^^- 1949 1950 -12- 50 T- 1 r 100 150 1^ X 1951 -?^ 1952 ^-£?- 1953 1954 -£",?- 1966 200 LENGTH (cm) ^■?-T 1 1 "1 1 250 300 50 100 150 Jt_ 1 200 -r — I £50 300 LENGTH (cm) Figure 30. Lengths of blucfin tuna captured from the Bahamas ("+" on graph means less than 5%). 32 Figure 31 Geographical references for the Gulf of Mexico and Caribbean Sea 1967a) Small areas of high catch lalcs also occurred about midway be- tween easternmost Soiiih America and westernmost Afhea In the third quarter, the only ar- eas of high catch rates were off New England and the Grand Banks of Newfoundland, and again in the nar- row part of the Atlantic, midway be- tween the most western part of Af- rica and the most eastern part of South America In the last quarter, the concen- tration south of the Canadian banks was farther west and extended south- ward to latitude 30°N. The concentra- tion between South America and Af- rica had moved northward to 20°- 32°N The concentration off eastern- most Brazil had formed again, and 4*- ' ' I ■ ■ I I I Figure 32 Distribution of oceanic catches of bluefin tuna (per 1(),(K)0 hcx)ks) in the lour quarters of the year, 1956-1968 33 r15 O NUMBER OF FISH/10'' ■ METRIC TONS/10* F ^ «■ P 1955 Figure 33. Annual catches of Japanese longline fisheries in the Atlantic Ocean for large and medium size bluefin tuna in tens of thousands of fish and in thousands of tons. extended northward to 1 6°N, almost joining the mid-ocean one men- tioned above. Considerable size data are avail- able from exploratory fishing and commercial catches in the western North Atlantic These records indi- cate that most of the bluefin taken in oceanic waters south of 35°N and west of70°W (catches north of 35°N and west of 40°W have been dis- cussed in a previous section) were giants. Unfortunately, little size data are available for the much more ex- tensive Japanese catches. Hayasi et al. (1970) reported that the Japa- nese longline fishery took "small" bluefin off Flonda (25°-35°N, 70°- 80°W), but no size data have been provided. Hayasi and Shingu (1972) and P. C. Wilson (personal commu- nicafion) provided lengths for about 220 Atlantic bluefin caught in oce- anic waters Most of these were gi- ants, but a considerable number were medium sized, although mostly of the larger sizes (ages 7 and 8) oc- curring in that group. It appears that nearly all of the bluefin over one year old taken in the westernmost part of the North Atlantic south of 35°N, and in the Gulf of Mexico and the Caribbean Sea, were giants, but that there have been considerable numbers of medium sized, as well as giant, bluefin in mid ocean catches in the 1960s (Shingu et al. 1975). After the decline of their Atlan- tic bluefin tuna catches in the late 1960s and early 1970s, the Japanese shifted much of their effort to the eastern Atlantic and the Mediterra- nean. The remaining effort in the western Atlantic was concentrated in the northern Gulf of Mexico in May and June, when giant bluefin spawn there, and in the area south of the Grand Banks through much of the year (Shingu et al. 1975). The move into the eastern At- lantic brought new pressure on stocks of larger fish which had previously been fished mainly by the declining trap and seine fisheries, and on me- dium sized fish which had been pre- viously fished mainly by the Spanish and French bait boats in the Bay of Biscay. The fishery also provided in- dications of a previously suspected wintering ground around the Canary Islands (Aloncle 1964) for the large bluefin which were fished in Euro- pean waters in the waim season. The effect of these shifts in ef- fort was to restore the catches to levels approaching those of the mid 1960s. Whether they will follow the pattern of rapid rise and rapid de- cline illustrated for other areas by the combination of Figures 34 and 35 remains to be seen. Shingu and Hisada (1976) showed the length compositions of samples from longline catches taken in waters west of Gibraltar in May- August 1973 and Apnl-June 1975. The former was dominated by large fish 1 90-240 cm long with a mode at 205-215 cm, but also included a significant number of medium sized fish 135-155 cm long. Larger fish, 185-300 cm long with a broad mode at 225-255 cm, were even more pre- ponderant in the latter sample, with only a scattering offish 1 10-175 cm long. 4. Eastern Atlantic Islands Bluefin tuna occur in the vicin- ity of the Azores, Madeira, and the Canary Islands (Figure 36), but there has been little information on the fisheries and the sizes of fish until recently. a. Azores Ferreira (1932) reported on the tuna fisheries of the Azores for the years 1924-31. The principal catch was bigeye tuna, " Paratlmnnus obesiis", with bluefin tuna being much less frequent. Seven of 219 tunas caught in 1930 and 5 of 1404 caught in 1931 were bluefin. The fishing seasons usually extended from mid-April or early May until July or August. Ferreira docs not state the sizes of fish taken, but mentions that very small (50 cm) tunas occurred, as well as adults. Figure 5 in his work, which is described as a young ''Parathitmnis obesus" of 51 cm, shows, in our opinion, a young blue- fin tuna. During cruise 63-4 of the Bu- reau of Commercial Fisheries M/V "Delaware", six giant bluefin were taken by longline off the Azorean Island of Santa Maria in May 1963, and another was lost alongside the 34 Tabic 8 Catch rates (numbers of fish per 1 ,0()0 hooks) for bluefin tuna of the Japiincsc Atlantic longhnc fishciy, 1 956-7 1 , by year and area The BEN and RIO areas arc not included since it is believed that the bluclin caught in them were predominantly southern bluefin. Note: 0.00 = O.005 but > 0, U = effort but no catch, - = no effort Year GM NOW NOE CAR GUI CV GG BAD Total 1956 0.01 . . 001 1957 - - 0.14 0.11 0.07 1958 - - - 0.01 0.01 0.00 1.17 0.06 1959 - 0.51 0.19 0.08 0.00 094 0,22 1960 - 0.81 0.15 0.17 0.32 0.33 1961 - 013 1.11 0.08 0.01 0.36 16 1962 - 0.02 0.13 0.01 3.12 0.36 0.21 083 l.Il 1963 0.02 018 0.05 003 3.94 0.25 0.01 3 82 1.42 1964 18 0.99 0.21 019 1.78 0.12 1,07 0.88 1965 2.98 16 0.07 0.76 0.07 0.00 0.26 0.76 1966 10 3.39 0.07 003 30 0.01 0.00 002 0.58 1967 0.72 0.03 0.11. 0.04 0.00 0.00 0.17 1968 0.31 0.34 0.02 0.01 0.02 0.00 0.00 0.08 1969 0.01 0.18 0.00 0.02 0.00 0.00 0.04 1970 0.00 0.04 0.06 0.02 0.00 0.02 0.02 1971 0.43 0.12 0.00 0.00 0.19 1956-71 TOTALS Thousands of Fish: 0.7 94,0 2.0 0.8 130.0 9.9 3.2 52.3 293.0 Millions of Hooks 8.9 87.5 17 3 196 96.4 91.2 92.6 58.4 471,8 Fish per Thousand Hooks: 0.08 1.07 12 0.04 1.35 0.11 0.03 089 0.62 vessel (Anonymous 1963) In the same period, six bluefin tuna were observed among the landings of the local fishery at Ponta Deigada, and all of these were also giants (obser- vations by Dr B B. CoIIette, Dr. D de Sylva, and F J. Mather). b. Madeira The numbers of bluefin tuna sold in the market in Funchal, Madeira, in the years 1954-1962 were furnished by J. Maul (personal communication) These ranged fi^om six to 268 fish per year, averaging 104. The most nu- merous landings were in March-June and in October, with the least in No- vember-January. Catches of other tu- nas, notably bigeye, were much more numerous. Maul has stated that the bluefin is relatively low in abundance off Madeira, and that the catches are from two size groups: 6-7 kg, and 1 50-200 kg (Aloncle 1966) c. Canary I.slands Early inlbrmation (Frade 1929) indicated that T. ihynnus thynnus was rather rare at the Canary Islands, but more recent developments indicate that the .species may occur there in considerable numbers, and in vari- ous sizes Aloncle (1964) reported a wintering area for small and young adult bluefin tuna (0.5-60 kg) be- tween Lanzerote, one of the eastern- most of the Canaries, and the Moroc- can coast fhe senior author directed the tagging of four 45 cm (age 0) fi.sh oil Gran Canaria in December 1974 Catches of large bluefin tuna ha\c been made by sport fishermen off ■fenerife m the Canaiy Islands, where the "European ' record was broken twice in two years, with 374 kg and 392 kg fish (Anonymous 1975,1 976). Several additional catches of large bluefin, off Gran Canaria, were re- ported by C H. Roncoroni (personal communications) It appears that most of these giants were taken in laic fall and winter (L F. de Gamboa, per- sonal communication). The distribu- tion of Japanese winter (January- March) longline catches in 1974 (Fisheries Agency of Japan 1976) in- dicates that the Canary Islands are in a wintering area for bluefin tuna. Santos (1 976, 1977a, 1977b) has provided the first detailed informa- tion on commercial catches of blue- fin tuna in the Canary Islands The 1975 landings, excluding longline catches, were 932 metric tons. Catches and catch per unit of effort were greatest in the period June-No- 35 Figure 34. Areas of the AUanlic Ocean (Wise and Davis 1973). vember, peaking in August and September In 1 976, howe\'er, Santos reported that catches were frequent in the first four months, but dechned during the summer. The estimated catch through September was 641 mt. L. F de Gamboa (per- sonal communication, November, 1976) m- formed us that the sport fishing off Tenenffe followed the same trend, and that the excellent fishing which had occurred in the fall of 1975 had not materialized in 1976 5. Eastern Atlantic Bluefin tuna fisheries in the eastern Atlantic (Figure 36) have been distributed in a manner somewhat similar to those in the western Atlan- tic. Large and medium-sized individuals have been taken during summer and early fall in Nor- wegian waters and the North Sea 'Ihc largest fishery for small bluefin, with some mediums, has been in the Bay of Biscay, mainly in sum- mer Less important fisheries for small bluefin occur off Portugal in the fall, and off Morocco during much of the year. 'I'rap fisheries in the Ibero-Moroccan Bay have taken important catches of large bluefin, along with usually lesser amounts of medium and small indniduals, in late spring and earh' summer The fisheries in Norwegian waters and the North Sea are of comparatively recent origin, as IS the one for small bluefin off Morocco. The Bay of Biscay fisheiy is over a centun' old, but was modemized about 1950 The trap fisheries in the Ibero-Moroccan Bay, on the other hand, arc of extremely ancient origin. The Japanese longline fisheiy has become active in the Bay of Biscay and the waters olY Gibraltar and northwestern Africa since 1 970 •- EFFORT WITH CATCH X - EFFORT WITHOUT CATCH GM X / NOW -i4 3 2 1 x-x — • -' X --Xi ^^~*~ « ' NOE j: g X X x - x — X - CAR X — — •-S ( X ]; / ■■-—->-• / \ GUI *-.•—•- -|4 -3 -2 -1 O 9-mm—»—^~»' -X-M GG ^; X X / •\.-.-- \ BAH I cv 1 ' — •-•— •-— ■— •—x^-Jq X-IQ / \ TOTAL - _l L _l 1 1 — i- -|3 -2 - 1 O 2 1 O 1956 60 65 70 Figure 35 Japanese longline catch rates of hiuefin tuna in areas of the Atkmtic Ocean (Note: Bluefin ui aieas RK) and BEN aie assumed to have been southern bluefin, Thunnus maccoyii. while those in oilier areas are assumed to have been Atlantic bluefin, Thunnus ihynnus ihynnus.) 36 20° 0' Figure 36. Geographic references in the eastern Atlantic a. Northeastern Atlantic Le Gall (1927, 1929) described the distribution of bluefin in the north- easlcm Atlantic and the North Sea (Figure 36) The species was re- corded in abundance off the south- west coast of Ireland in July and off the north coast in July and August They were also observed off the west coast of Scotland in July, and off the south and west coasts of Norway, and in the Skagerrak and the Kattegat from July to October Hemng boats encountered them in July and Au- gust east and south of Fair Isle, then north and west of Dogger Bank In late September and October, bluefin were observed on the west edge of Dogger Bank In November and De- cember, trawlers on the Great Sole at the edge of the continental shelf en- countered schools of them surfacing l.e Gall related the occurrences of bluefin in the North Sea and its tribu- taries to the seasonal "transgressions" of Atlantic and Atlantic slope wa- ters, with salinities of about 35 o/oo b. Norway Bluefin tuna have been taken by harpoon and angling gear in Norwe- gian waters for many years, but the introduction of the purse seine method in the late 1940s greatly in- creased the catch The fishen', includ- ing the catch and its size composi- tion (Figure 37), has been described by I lamre ( 1 97 1 ). The catch increased from a few hundred tons in the 1 940s to over 1 1 ,000 tons in 1952, but de- creased after 1962 to 2,500 tons or less per year. Hamre attributed the decrease to lack of recruitment, lie showed that the seasonal movement of bluefin tuna through the fishery varied with age of fish. The largest fish (age 7 and older) arrived off Bergen in early July and migrated to the northern Norwegian coast. After feeding there for three or four weeks, they migrated southward into the bank area of the North Sea Fish of ages 6-12 arrived in mid- July south of 62°N latitude, and migrated south- ward along the coast. Catches of fish of this size in late autumn by Swed- ish and Danish fishermen in the Kattegat, including one fish which had been tagged earlier in the same season in Norwegian waters, indi- cated that the migration continued to that area Five and six year old blue- fin arrived in the southern area in September and migrated to the east coast of Norway (Skagerrak). I'his general pattern was fol- lowed as long as new vear classes were recruited. Since 1958, when the 1952 year class was recruited, there has been no recruitment of younger age groups to the stock (Figures 37- 39) This resulted in a decline in the annual catch, and a change in the migration pattern of the fi.sh after the 1 962 season. The largest bluefin now rarely migrate to the northward, but usually follow the southerly route pre- viously travelled by the intermediate size (6-12 year old) fish A fev/ large fish, however, were taken in the northern area in 1967-1969 (Figure 38) No bluefin have been caught in the German North Sea fishery since 1 962, probably because of this change 111 migratoiy pattern, but a few have been caught in the Kattegat. Medium sized bluefin have failed to appear in the Norwegian fishery since 1962 (Figures 37-39) The annual catch since 1 967 has been less than 1 ,000 tons. It fell to a low of 90 tons in 1972 and even less in 1973. An in- crease in the modal and minimum sizes of fish taken since 1 957 is evi- dent for botli the northern and south- 37 em Norwegian areas, but is espe- cially striking for the soutiicm area (Figure 39) c. North Sea The German bluefin tuna fish- ery in the North Sea was carried out by handline, with the capture some- times facilitated by an electrocuting system (Meyer- Waarden 1951) The landings reached a maximum of 1,286 tons in 1957 but declined to 194 tons m 1962 (Tiews 1975). The fishery was then abandoned because of lack of fish. The catch consisted entirely of giant fish (Figure 40) An increase in the modal size of the fish taken since 1952 is apparent A simi- lar Danish fishery' took catches of from 800 to 2,100 tons in the years 1950-1955, but became negligible after 1959 (Tiews 1975). Small catches of bluefin tuna ha\'e been taken in the approaches to the Baltic by Danish fishermen using various gears Yearly catches since 1960 have been less than 200 tons. The fish taken have been large, with the modal size increasing in recent years. Sport fishing in the Oresund near Elsinore, Denmark, in 1948- 1954 and 1960 produced yearly catches of from 18 to 119 fish with average weights of from 120 to 260 kg (L. R. Crandall, personal commu- nication). English sport fishermen have oc- casionally taken giant bluefin tuna m the North Sea. The British 'funny Club (Anonymous 1937) reported that tuna appear off the Shetland Is- lands in June and between Scarborough and the Dogger Bank off the Yorkshire coast from July to October. Sport catches in the latter area from 1932 to 1936 were listed as 21, 80, 54, 53, and 33, respec- tively. The veiy difficult conditions for this sport, combined with the lack offish in recent years, have prevented its growlh. d. Bay of Biscay The most imporlaiU fislicry for small (2.5-35 kg) bluefin tuna in the eastern North Atlantic has been in the Bay of Biscay (Figure 36). fhis Bay is therefore, prcsuiiiubly, the major nursery ground for yi)ung blue- fin in the region and is a major source of recruitment to the fisheries for mc- D NORTHERN AREA SOUTHERN AREA No. X 10 50-1 25- J 25- O-J 0-^ 25- 25- ?5-i % of Sample - ^^- i \ s- / \ ~N 1956 1 / \ -• X \ - / v. y^ * V ^..^ f\ / U -= 10- /' / r \ / \ . \ 5^ / \ / \ \ 1957 / \ 1 \ / ^ ^^^ *-'s. ' ^ 1 \ ,^^ 0" ,' s ■v 1 > / \ 1 V 5-^ / \ / \ 1958 1 iw^^^ 1 \ ■v/ N ^^^ /*^ .^^ V^ n / ^X ^^~ ^ ^^ 10- / "» / \ / \ 5-; /-, r 1959 1 X X /■ 1 N \ / / - Z. ^. y ^>i;»^ 0— ' 1 / '-N 5-: / -^ / V \ I960 \ / \ X / • \ •».. Z ^-r -> — ^^ ^'^ — /~ \,. •-•v / /s i 19S1 5 — f /, s. V / • \ \ ^ / \ \ y ■^ / N^^ y^ ^f — ^ /^ v" « J X \ / / A 1962 5 — / 1 N \ / 1 ■v \ / / \ \ ^ v,-^._ — /" ■'X^ 1963 5 — / / _y -^ ^^--^^^ — / "X '"\ 1964 5- ^/ \^ ^ J 1 ^s^^\ / / \\ r^ ,/ ^•-*w -_ 0- / / >v 1965 5-^ r^ / / 0- 1 Ij > ' I n-i rr I 1 I 1 1 1 I I I 1 1 1 1 1 1 • • WEIt iHT: 10 D 2 00 300 kg AGE: 6 7 8 9 1G 11 12 13 years Figure 37 Weight distnbution ol bluefin tiuiu ^.apiured from Norwegian waters bv aiea and yemv llic columns to the left show iiumbci of fish landed with one unit equaling 5,000 fish (from Mamre 1 97 1 ). 38 20 SO 20 10 30 20 10 SO 20 iO - \95B -a- 1937 1968 -a- 30 r «59 10 - so 20 - 10 1960 **- SO 100 «96t -a- 1967 1966 -a- 1969 ■»T r 1 r ^^ 50 100 150 ZOO 250 300 LENGTH (cm) LENGTH Figure 38. Lengths of bluefin tuna captured off Norway (north of 63°N) ("+" on graph means less than 0.5%). dium sized and large bluefin in the eastern Atlantic Considerable num- bers of medium sized (32-122 kg) and some large bluefin have also been taken in the Bay of Biscay (Le Gall 1954, Shinguet al 1975) French tuna landings from this bay began about 1840, when pilot boats from La Rochelle began taking considerable catches of tuna by tioll- ing while waiting for ships Tuna fish- ing did not become important there commercially, however, until a crisis in the sardine fishery about 1860 caused the French fishermen to turn their attention to tuna. The success of the trials exceeded expectations, and the fleet of sailing trollers increased year by year (Grandbesan(;on 1 909) The major catch of this fishery was albacore, Thunnus alalunga, with bluefin usually being taken inciden- tally. The introduction of the live bait method in Spain and France in the years 1947-1949 (Navaz 1950a, 1950b, dc la Tourrasse 1951), how- c\cr, resulted m a specialized fishery for bluefin The catches of this spe- cies consequently increased greatly in the 1950s, but were much lower through most of the 1 960s and have remained at intermediate levels in the 1970s (Figure 41) Statistics for this fishery are VC17 confusing, but it ap- pears that the landings varied between 1,000 and 1.500 tons m 1945-1949, then rose to between 2,700 and 5,500 tons in 1949-1959. The catches in the 1960s were much smaller, gener- ally varying between 1 ,000 and 1 ,900 tons, but attained 2,100 and 3,300 tons in 1965 and 1966, respectively. Catches in the early 1 970s have evi- dently been somewhat over 2,000 tons per year. Bluefin tuna apparently occur in the Bay of Biscay in every month of the year (J. LeGall 1950. 1954; Navaz 1950b), but the active fishing season has usually extended from May or early June into October or November Individuals weighing more than 30 kg are usually taken between mid-July and early Septem- ber (llamre and Tiews 1964, Hamre et al. 1966, 1968, 1971, Aloncle et al, 1974), Research on bluefin tuna m the Bay of Biscay has been divided be- tween two periods — 1 949-1 954 and from 1972 to the present This divi- sion of research effort leaves a gap of 39 30 20 - 10 - 30 20 10 o 30 20 10 «955 -tf- 1956 19S7 O ^i~ 30 r 20 10 o 30 20 Vi fi 10 I 5 ° ^ '° J» 20 10 o 30 20 10 o so 20 10 o 30 195B ■^r- 1959 '—a- 1960 1961 1962 20 \- 1963 10 30 20 10 o 1964 -a-r 150 200 L£NGTH (cm ) 1969 ^^- 1966 4967 -If- 196B 1969 -£Z- 1970 ■tf- 1971 1972 V.?- 1973 1974 3O0 50 i^ i. 150 200 Z50 300 LENGTH (cm) Figure 39. Lengths of bluefin tuna captured off Norway (south of 63''N) ("+" on graph means less than 0.5%). 40 30 - 20 1951 10 30 - ♦ •♦* _^^^^^^^^^^^^k 20 1952 10 30 - 20 1953 5j 10 5 r 20 1954 10 30 W^ 20 1959 10 rr 30 20 1956 10 -*^T . T 1 T 1 — ^"^ 1 1957 1938 ^f- 1999 -«- I960 1961 -rr- 1962 50 100 JO -i 1 — 100 >-!l^ — I r 150 -T^^-^f^ 1 200 250 300 LENSTHtcml LENGTHkml Figure 40. Lengths for bluefin tuna captured from the Noilh Sea ("+" on graph means less than 0.5%) almost two decades for which ver)' few data are available. Length frequency data in 5 cm groups for Bay of Biscay catches in 1949 (J Le Gall 1950) (.French catches only) and 1972 (Bard et al 1973) and 1975 (Cort 1 976) (for both French and Spanish catches) are pre- sented in Figure 42. Additional size frequency data, in different size groups, are furnished by Dao and Bessineton (1974) and Cort and Cendrero (1975) These data indicate that age 2 fish are usually the most important in the Bay of Biscay catches, with age 3 the next in order, and age 1 fish oc- curring episodically in recent years (Dao and Bessineton l')74). Age 1 fish were important in the 1949 sample, but no data for other years in the peak period of the live-bait fish- ery are available There has been a decrease in the importance of age 4 and larger fish in the catches since 1972 This has been attributed to the entry of Japanese longline vessels into the area in 1 974 and subsequent sea- sons (Cort and Cendrcio 1975, Cort 1976), Much additional data is avail- able for weekly (or daily) landings at St. Jean de Luz, the principal port of the French bluefin tuna fishery, in terms of fish weighmg less than 30 kg and more than 30 kg (Hamre and Tiews 1964, Hamre etal 1966, 1968, 1971, Aloncle et al. 1974). Eighty percent of the 1962-1972 landings by weight were in the class weighing less than 30 kg (Dao and Bessineton 1974). The percentage by numbers would, of course, be considerably higher The Spanish fishermen, how- ever, take more of the larger fish than the French Dao and Bessineton (1974) found that only 52% of the total French and Spanish catches of 1972 and 1973 consisted of fish weighing less than 30 kg, whereas 67% and 80%, respectively, of the French catches for those years were offish weighing less than 30 kg. They found two apparent causes for this difference — the French sought out the smaller fish, which command a higher price, — the baits used by tire Spaniards were larger than those used bv the French. In view of the virtual disappear- ance of medium sized bluefin tuna from the landings of the Nonvcgian fishery since 1962, and their rela- tively poor showing in the trap fish- eries of the Ibcro-Moroccan Bay (sub- part g of this section), it is unfortu- nate that size composition data for Spanish landings prior to 1 972 are not available. In addition to the commercial fisheries, there has been an active sport fisher}^ along the north coast of Spain for bluefin tuna and albacorc. Reportedly (L. F de Gamboa, M R. Borrell, personal communications) this fishery has declined greatly in recent years because of lack of fish- ing success In addition, giant blue- fin have been taken by Michael and Helen Lerner in August 1947 off Trehurden (Normandy, France) (Famngton 1949) and in the Bay of Biscay by the late Cleneralissimo Franco and Max Borrell (Max Borrell. personal communication). e. West Coast of Portugal Small bluefin w ere taken by troll- ing in the vicinity of Cape Espichel (south of Lisbon), Portugal, in late summer and early fall. The season usually started m September-Octo- ber (Vilcia and Monteiro 1961, quoted in fiews 1963) In 1960 the entire catchof5,500 fish (34 65 tons), awraging 6 3 kg, was taken m No- vember The 1961 catch was 7,859 fish totalling 36.01 tons (4 6 kg aver- age) (Hamre and Tiews 1964) Both years" catches consisted of a large group of age 1 fish (50-70 cm) and a smaller group of age 2 fish (75-85 cm), as shown by histograms for each year and a sample of 128 fish from September 1961 catches presented by Hamre and I'lews (1964). The 1965 catch is illustrated by the length hequencies of a 363 -fish sample tabulated by week of capture (weeks 40-46, September 26-Novem- ber 1 3), All of the 363 fish measured were age 1; their total weight was 2,047 kg and the average was 5 6 kg The total catch of about 13,000 fish weighing 75 tons was taken mainly between early October and early No- vember The 1966 catch was small and irregular and was not recorded (Hamre el al. 1968) In l')68, 26,199 fish averaging 5 kg were taken (Hamre ct al. 197] ) f. West coast of Morocco Small bluelin are taken olT the west coast of Morocco hy hook and line, live bail, and .seine fisheries (ihc trap fisheiy will be discussed sepa- rately). Catches ranged up to 2,000 tons in the mid 1 960s, but have been less than 1000 tons in recent vears (Figure 43). Fish are present during most of the year, but the largest catches occur in autumn (Aloncle 1964) Aloncle (1966) gave the length composition of a 9 1 -fish sample from a purse seine catch of 700 kg of blue- fin taken September 9, 1967, at 30M0'N, 10°05-W (off Cape Ghir). The lengths ranged from 50 to 67 cm, with nearly all tlie samples in a modal group extending from 56 to 65 cm, and having its main peak at 62 cm. This indicates that nearly all the fish were age 1 Aloncle stated that bluefin of this age were common in the region at this season. I'his catch was made duung a .series of experi- mental cruises of the seiner "Danguy,"" which extended from Sep- tember 1964 to July 1965, and cov- ered the area bounded by the African coast from Tangier to Cape Bojador, the south coasts of Spain and Portu- gal, the Canary Islands, and the Ma- deira archipelago I'he first lew weeks of the cruise resulted in the capture ol age and age 1 bluefin and sightings of age 2 and possibly age 3 bluefin The few bluefin taken by trolling during the winter were in the 62-cm class In late April and early Ma\' two indi- viduals, one 70 em and one 4 v5 cm long, were taken northeast of the Sal- vage Islands These wouki ha\e been ages 2 and 1 , respectiveh . in the en- suing summer In Novcmbei and December 1960, fishing boats from Barbate, Spain, seined a great quantity of blue- fin about 42 cm long and 1.7 kg in weight, along with young albacore of about the same size (Rodriguez- Roda 1964a) A histogram of a sample of 100 of these showed a range from 38 to 45 cm, with a mode at 419 cm. This author slates that the fishermen of Barbate often .seine blue- fin of 40 to 60 cm in length off the Moroccan coast from Larache to Casablanca and even to Safi and Agadir in October, November and December g. Ibcro-Moroccan Bay Trap Fisheries As noted previously, trap fisher- ies for bluefin tuna have existed in the Ibero-Moroccan Bay (Figures 44 and 45) for centuries. The recorded histories of the Spanish and Portu- guese fisheries date back to the 1 5th and 16th centuries, respectively (Pavesi 1889) On the other hand, we have found no records of Moroccan Atlantic traps prior to the 20th cen- tury The Phoenicians and Carthaginians who fished intensively for bluefin in all these areas in the pre-Christian era (Parona 1919), how- ever, probably operated traps similar to those still in use (Thomazi 1947). The specialized tuna traps are very large and complex structures, with a leader extending up to 5 km, and sometimes e\en more, from the shore to the bod\' of the trap. In many cases an additional leader extends diagonally up to 2 km farther off- shore (Figure 46) Traps have been described and illustrated by several authors (Pavesi 1889, Parona 1919, Rodriguez-Roda 1964a, Sara 1964, de Cristofaro 1970). Two basic types of tuna traps have been used exten- sively, the "Atlantic" or "Spanish" trap and the "Mediterranean" or "Si- cilian" trap Fodera (1964) and de Cristofaro (1970) described the de- sign, construction and relative advan- tages and disadvantages of each r\pe. fhe traps in the Ibero-Moroccan Bay were of the Atlantic type Most of the traps along the Span- ish and Portuguese coasts faced west and fished the "arrival" (eastward) 42 O NUMBER OF FISH/10» ■ METRIC TONS/10' Figure 41 Annual catches by Bay of Biscay fisheries for small and some medium size bluefin tuna in hundreds of thousands of fish and in tlioasands of tons. run from late April through June. Some of these were reversed at the end of June and fished the "return" (westward) run in July and August. The Moroccan traps faced southwaid and fished the "arrival" run, which was northward in that locality, only After the 1976 season it appeared that, of all these traps, only one or two located off the Spanish coast would continue to operate These fisheries, and the biology of the fish which supply them, have been the subject of a great many in- vestigations and scientific papers. There are two maxima in the fishery, one near the middle of the "arrival" run, and the other near the middle of the "return" run Spawn- ing is presumed to occur between these maxima (Sella 1929a, 1929b; Vilela 1960) Sella (1929b) provided generalized diagrams showing the variation of the catches with dates and the movements of the fish in the vicinity of the traps during the spawn- ing cycle (Figure 47) For fish of the same length, those taken in the "arrival" run average about 15% heavier than those taken in the "return" run (Rodriguez-Roda 1964b) (data from fish taken at Barbate, 1956-1961) Vilela el al (1960) found that females lost about 21 percent of their weight between the two runs, when most of the spawn- ing takes place, while males lo.st only about 10 percent of their weight Length frequency data for the Spanish fishery for the vears 1956- 1959, 1961 and 1963-1975 are pre- sented (Figure 48) 'fhe catch con- sisted mainly of medium and giant 30 r fish, with small ones (ages 1-2) significant in 1958 and 1959 only. Medium sized bluefin were domi- nant in 1957 and important in 1 956, 1 958, and 1 959. Since 1 960. however, giants have constituted over 70 percent of each year's sample One year class, probably that of 1954, dominated the samples for the years 1963-1965 (Rodriguez-Roda 1969b). Length frequency data are not available for the Portuguese trap catches, but the yearly catches for the years 1931-1972 (Vilela and Cadima 1961, Hamre and Tiews 1964, Hamre et al 1966, 1968, 1971) are shown in the size groupings traditionally used in this fishery (Figure 49) Fish of the "atuns" size were usually dominant, but in three years the cachorretas, nearlv the same sizes as our "small" group, considerably outnumbered them. The.se fish were generally most 20 - 10 - 30 7S o 20 - 10 - 30 20 - 10 150 Length (cm) 200 250 Figure 42, Lengths of bluefin tuna captiued fi-om the Bay of Biscay ("+" on graph means less than 5'Mi) 43 □ number OF FISH/10' ■ METRIC T0HS/'1O> 1960 1-4 - 3 - 2 . 1 Year Figure 43 Number and weight of fish tor Ihc Morocco fishcnc medium size bluefin tuna. for small and numerous m late August (Vilcla 1960) The albacoras and atuarros (ages 4-7, mcluding most of our "me- dium" group) were important in some years but, as has happened with the medium group in most other areas, have generally decreased in impor- tance in the most recent years Size composition data for Span- ish trap catches by week (Rodriguez- Roda 1964b) showed a tendency for the largest individuals to be taken near the beginning of each "run." Aloncle (1964) observed the same tendency m the catches of the Mo- roccan tiaps. Similarly, Vilela (1960) showed that the smaller tunas tended to be most important in the catches toward the end of each iiin, but espe- cially in the "return" run. Vilela (1960) found 300 males (38.1%) and 487 females (61.9%) m a sample of the bluefin taken in the Portuguese fishery in the years 1958- 1960. There was little year-to-year variation from this proportion There was also little difference between the sex ratios for the "aiTival" and "re- turn" runs, except in 1960, when the sample from the "arrival" run con- sisted of 44 4%) males and 55.6% fe- males. Rodrigucz-Roda (1964b) pre- sented the figures for the sexes deter- mined from samples of the 1 956- 1 96 1 catches of the Barbate trap, and the 1961 catches of two others (Tabic 9) Lozano Cabo ( 1 958) pointed out that the average size of the bluefin caught in the Portuguese and Span- ish traps increased according to ho\\ far east (near Gibraltar) the traps were located The average was smallest m the Portuguese traps, and the greatest in the trap at Tarifa, near tiihraltar He reported that the average si/.e of the bluefin taken in the Moroccan traps was even greater than the aver- age of those taken at Tarifa. Aloncle (1964) showed that in Morocco the average weight of the fish caught also increased with the proximity of the traps to Gibraltar The largest fish were taken at Cape Spartel. at the entrance to the Strait and opposite Tarifa Statistics for most of the Ibero- Moroccan Bay trap catch are shown m Figure 50 These do not include all of the Moroccan catch We were unable to obtain continuous data for the important trap at Cape Spartel (in the former International zone of Tangier) for 1933-1953, or for the three more ephemeral traps near Kenitra (in the former French Pro- tectorate) for 1939-1955. Therefore we have omitted their catches and used those of the Larache group (in the fonncr Spanish Protectorate) to represent the Moroccan catch. The catches of this group, which varied from one to five traps in the years 1927-1954 and stabilized at three from 1 955- 1 966. have been recorded, in numbers of fish, for 1927-1962 (Lozano Cabo 1958, Hamre et al. 1 966) Data for subsequent years are available, in various forms, in Aloncle ( 1 964), Collignon ( 1 964, 1 965, 1 966, 1967, 1968, 1969, 1972), Lambeouf (1 972), and personal communicafions fromM l.amboeuf and R. Sara Since 1 966 only one or two of the Larache traps, if any, have been set. Data for » CITIES 3> GCOCRAPMICAL RCFIRENCES 1. C.SANTA MAMA 2. CTfUFALOAII 3. CSPAinTL 4. tTIIAIT or aa*LTAR 5. CltRAUAR C. CEUTA, R ALMINA . ;', 7. GULF OF VILC2 y^^'^ a ALSORAN II 9. C. Ties FDRCAS y WESTERN MEDITERMANEAN SE* SC* OF ALIOAAN) •(g 2" Figure 44, Geographic references for the lbero-Morocc;in Ba\' area. 44 at'- CStVINCINT -^iMiitaiisii^^jiUiiiB 10* fCtTIES TRAP SYMBOLS VC«R L*>T RcroRTCO StT KRIOO S FISHCO | ARRIVAL RETURN BOTH 1920 1949 o a A t«Sa'1971 ® H A 1972 • ■ ▲ Figure 45. Trap fishing locations for bluefin tuna in the Ibcro-Moroccan Bay area. the Spanish traps, provided by Lozano Cabo (1958), Rodriguez- Roda (1964a, 1973, 1974), and Aloncle et al (1976), are believed to be complete, but are somewhat bi- ased by the inclusion of some records of the La Linea trap, which is actu- ally just inside the Mediterranean. Its catches, however, were relatively small From four to seven Spanish "Atlantic" traps were listed for 1 94 1 - 1971. Since then, only one or two, if any, have been set. The Portuguese data, provided by Vilela (1960), Lima Dias and Barraca (1972) and Republica Portuguesa (1957, 1958. Figure 46. Schematic diagram of the type of sfx;cializcd tuna trap used in the Ibero-Moroccan Bay area: (a) body of trap, divided into compartments, (b) leader fi"om shore to trap, (c) leader extending olTshore from trap Solid hnes show trap arranged to catch fish traveling from right lo left Dashed lines show trap modified to catch fish traveling from left lo right. 1959, 1960, 1961, 1962, 1963, 1964, 1965, 1966, 1967, 1968, 1969, 1970, 1971) are believed to be complete. Five traps were set in most of the years 1931-1965. three m 1966-1968 and two in 1968-1972. Due to these discrepancies and ormssions, the data m Figure SO are not exact. They are presented to illustrate trends in the catches rather than exact totals. Pavesi (1889) estimated the mean annual production of the Span- ish traps for the penod 1884-1887 at 70,000 tuna, or quintals (100 kg). He derived a corresponding figure of 30,000 tuna for the Portuguese traps. Roule (1917) calculated the average yearly catch of the Portuguese traps in the years 1896- 191 2 at 43,983 tuna. In the period from 1930 through 1960, the Spanish catches occasion- ally surpas.sed Pavesis figures, with 106,000 fish in 1930, 76,200 in 1943, 63,500 in 1944 and 80,500 in 1949. The average for the period, about 55,000 fish per year, was consider- ably below Pavesi's figure. The Por- tuguese traps only attained Pavesi's average of 30,000 fish in 1943, with 32,400 fi.sh, and never approached the 1896-1912 average of nearly 44,000 fish. The 1931-1960 average of the Portuguese fishery' was 17,400 fish per year The Larache group in Morocco produced a peak catch of 29,000 fish in 1958 and averaged 13,500 fish per year for 1930-1960. The addition of the Cape Spartel and Kenitra catches would increase these figures considerably Data for Mo- roccan catches before 1930 are too meager to permit any deductions in regard to long-term trends. The total annual catches of the samples of the three fisheries re- mained above 48,000 fish and 7,000 tons per year through 1962. A de- cline which began in 1963 became disastrous after 1967. listimated catches varied from 16,00(J to 24,000 fi.sh(2,400-4,300tons) in 1963-1967, and fell to 700 to 12,600 fish (100 to 1,600 tons) in 1968-1973. The Por- tuguese traps caught just one fish in 1971 (Lima Dias and Barraca 1972) and only 176 kg in 1972 (Y F. Barraca, personal communication). They have not been set since. In 1972, the only Spanish trap set in the At- lantic, Barbate, which had averaged 45 about 20,000 fish per year from 1 946- 1961 (Sakagawa and Coan 1974), took only 388 fish and the company which had operated the Spanish traps since 1929 was dissolved (Rodriguez Roda 1973) Barbate took 1,952 tuna m 1973 (Rodriguez Roda 1974). Two Moroccan traps, the only others set in the Atlantic in that year, caught 12 fish (R. Sara, personal com- munication). Two Moroccan traps were also set in 1 974, but took only seven bluefin (M Lambt)euf per- sonal communication). The situation had become such that when Baibate, which had not been set in 1974, took 1,842 fish (less than one tenth of the 1946-1961 average) in 1975, it was regarded as encouraging (Aioncle et al. 1976)! The decline in the Barbate catches was accompanied by a marked increase in the average weight, indicating that poor recruit- ment to the fishen,' was a major cause of the decreased catches Although the annual catch fell from 1 9,000 fish in 1 961 to 2,500 in 1 97 1 , the average weight per fish increased from 145 kg to 223 kg. (Sakagawa and Coan 1974). Rodriguez-Roda (1964a) had already noted the drastic decrease in the numbers of the smaller bluefin taken in the Spanish traps since 1953, and attributed it to either a high mor- tality on young fish in previous years, or unknown variations in oceano- graphic conditions which might have caused them to go elsewhere h. Trends in Eastern Atlantic Fisheries The decline in the northeastern Atlantic fisheries for large and me- dium bluefin tuna is illustrated in Fig- ure 51 This trend has been re\ersed to some extent by improved catches off Norway, and more significantly by the entry of the Japanese longline fisher}' into the area in 1971 (Shingu and Hisada 1976) Concern has been expressed o\er hea\y calchcs of \ci\ young bluefin as a possible cause for the decline in the fisheries for larger individuals (Rodriguez-Roda 1964a, 1964b, 1969d) 6. Mediterranean and Black Seas The fixed trap has been the prime producer of bluefin tuna in the Medi- terranean Sea (Figures 52 and 53) Table 9 Sexes dcteiTnined liom samples of 111 e 1956-1961 catches from the Barbate trap, ;uid the 1 90 1 calchcs of two others Trap Barbate Sancti-Pctri Isia Cristina Males Females 403 756 123 215 46 66 for centuries. The most important traps were the large ones in the cen- tral MediteiTanean. Ihe majority of these fished the "anival" run in May and June, the remainder fished the "return" run in July and August Some of the smaller traps also fished these migiatopj' passages and similar runs m the Bosphorus between the Sea of Marmara and the Black Sea The ma- jority of the smaller traps, howeser. fished mainh' for bluefin which were too young to spawn, or which were in the feeding, rather than the spawn- ing, phase of their annual cycle. I'herefore the fishing seasons of these traps were not limited to the May- August spawning cycle, and extended through much of the year. These smaller traps for non-spawnmg tuna were widely distributed along the coasts of the Mediterranean, but were PERIOD OF REPRODUCTIVE BEHAVIOR EMISSION & FERTILIZATION ARRIVAL IN THE ZONE OF MATURATION I DEPARTURE 100 -SO MAY a JUNE JULY A AUGUST ^^NING RETURN TRAP JULY - AUGUST Figure 47 Schematic tiiagram of bluefin luna niowmcnts in the vicinity of traps during tlie spawning cvcle 46 JOO 90 100 ISO zoo LENGTH (cm) Figure 48. Lengths of bluefin tuna captured in the Spanish lishcn y 1 956- 1959, 1 96 1 , 1 963 - 1 975) ("+" on graph means less than 0.5%). 47 50 40 30 I (O 20 10 N=NUMBER OF TRAPS B CACHORRETAS < 30 KG n ALBACORAS 30-49KG ATUARROS 50-89 KG ATUNS >90KG 50 31 32 33 34 35 36 37 38 39 40 41 42 4J44%5 46 47 48 49 50 61 52 53 5495 56 57M 5^^ YEAR Figure 49 Annual catches of the Portuguese trap fishery (Hamre iind Ticws 1964, Hamrc ot al 1966, 1968, 1971) most concentrated off France and what is now Yugoslavia (Pavesi 1889, Parona 1919, Belloc 1961). The history of the Mediterranean tuna U'aps dates to the pre-Chiistian era. Thomazi (1947) believed that the Phoenicians introduced this method throughout their colonics, and that some of the French traps had probably operated continuously from pre-Christian times until their demise near the end of the 20th century He thought that the Sicilian traps were of equally early origin, but had fallen into disuse during the Arab occupa- tion The Normans revived the fish- ery soon after their conquest ot' the island toward the end of the 1 2th century (Pavesi 1889, Parona 1919, Thomazi 1947). The numbers of tuna traps operating in the Mediterranean declined greatly during the 19th niui 20th eenUiries (Pavesi 1 889, Parona 1919), and especially since 1950 (de Cristofaro 1970). Until recently, the inaior tuna traps in the central Mediterranean have been of the "Sicilian" type Since World War II. however, chang- ing conditions ha\'e caused the aban- donment of many traps and the con- version of others to the "Spanish"' type, usually with modifications (Gaudilliere 1954, Fodera 1964, Sara 1964, de Cristofaro 1970). Various types of nets, hook and line gears, and haipoons ha\e also been used for bluefm tuna in many parts of the Mediteiranean-Black Sea system (Pavesi 1889, Parona 1919, Doumenge 195."^, lyigungor 1957). Until recently, however, their catches were small in comparison with those of the traps. Developments after World War II have altered this situation drasti- cally While the trap catches have declined catastrophically (Sara 1973), catches of small (I'ilic 1954, Scaccini and Biancalana 1959, di Meglio 1962) and large (Paini 1975, Mivake 1 976) biuefin by purse seine ha\e become important Also, Japanese longliners took increasing quantities of biuefin in the Mediterranean m the \ears 1972-1974 The catch de- clined in 1975, when the .hipanese government prohibited their longline vessels from fishing in the Mediter- ranean during the spawning season as a conser\'alion measure Meanwhile, promising sport fish- eries for biuefin tuna ha\e developed along the French and Spanish coasts from the mouth of the Rhone to Castellon and along the French and Italian Rivieras (Ligurian Sea'), where albacore arc also taken iCiianelli 1969, Cesarco 1972. M R. Horrell. A. Cesarco, I. F. dc Gamboa. and H K llarn-, personal comniunicalums') a. Western Mediterranean Although located between the fonneriy important trap fisheries of the Ibero-Moroccan Bay and the cen- tral Mediterranean, the biuefin tuna fisheries of the western Mediterra- nean ha\'c produced modest tonnages (Miyake et al. 1976). consisting mainly of small and medium fish Various gears have been used, some of which are of ancient origin. Oth- ers have been introduced recently. Among the most ancient are the traps, which were formerly wide-spread but now survive in a few locations off Spain and Africa only (Parona 1919, Belloc 1961) Hook and line gears and specialized nets have also been in use since long ago Purse seine and longline fishing have been intro- duced in tlic area since World War II. Sport fishing has also developed there in this same period 'I'he iTcnch traps are of histori- cal interest only, Thomazi (1947) slated that they reached their apogee in the 17ih century Gouiret (1894) reported those traps near Marseille took biuefin tuna from late July to late November, in 1851 there were 10 traps in the area, but in 1891 only three were operating. Their catch of biuefin in 1891 and 1892 (1,500- 2,000 fish, 41-4."! tons) was about hall' of the total catch in the same area bv other gears Parona (1919) 48 15-1 10- 5- P0RTU6AL- FISHERIES D NUMBER OF FISH /lO* ■ METRIC TONS/10' iiwihrtJrihmifrtwiftww^ ri5 -10 -5 1930 35 4b 45 50 55 ' ' 60 ' " 65 70 ° '5t SPAIN -fisheries "-is 10' 5- I IlllllillUldU 55 60 65 70 ■to -5 5-1 1930 35 40 45 50 55 60 65 MOROCCO -FISHERIES -D DATA FOR TRAPS OF LARACHE GROUP ONLY pnnpnnRnn, nl10n[1pnn[]f1f]f1n 1926 30 35 40 45 50 55 60 65 70 Voar Figure 50. Annual catches of Ibero-Moroccan Bay trap fisheries lor blucfin tuna in lens of thousands of fish and in thousands of tons. showed locations for 23 traps, of which eight were active, along the coast of Provence (France) between the mouth of the Rhone and Monaco Thomazi (1947), however, reported that the last of the French traps dis- appeared between 1 892 and 1 900 The Spanish and Moroccan Mediterranean traps have taken only small catches of blucfin tuna in re- cent years Parona (1919) showed 23 trap locations, of which nine were active, along the Spanish coast east of La Lmea. Belloc (1961) listed five in that area, with maximum annual catches of 60 tons of all species com- bined for a single trap m the years 1954-1958. In 1975 only one of these traps was set, and it took no bluefin tuna (Rey el al, 1977), Most of these traps fished from early February to mid October, and took feeding (non- spawning) tuna Other species were more important in their catches (Belloc 1961) Return traps have been set at La Lincu, on the Spanish coast near C'libraltar; at Ceuta, a Spanish city on the opposite side of the Strait of (libraltar, and at three nearby loca- tions on the Mediterranean coast of Morocco (Belloc 1961, Sara 1964) These traps take blucfin in .kily and August, but again they depend mainly on other species Yearly catches of bluefm tuna at La Lmea since 1961 have ranged from in 1 968 up to 2,400 fish weigh- ing 340 tons in 1965 Their average weights have varied from 128 to 186 kg, probably excluding the numer- ous age (1 kg) fish which are taken in some years (Rodriguez-Roda 1964a, 1964b, 1969d, 1973; Rey et al. 1977). The trap at Ceuta and those along the Medilcnanean coast of Morocco may bo considered together Only one or two of the Moroccan traps have fished in most recent years The blue- fin tuna catches have ranged up to a maximum of 1 72 tons for a trap, but have usually been 50 tons or less 49 (Collignon 1964, 1965, 1966, 1967, 1968, 1969, 1972, Lambocuf 1972, Crespo and Rey 1976, M. R. Horrell, personal communication) Little information is available on the sizes of bluefm taken. F. de Buen (1927) stated that 4-5 kg bluefin were taken in the Ceuta trap in early winter. Rodriguez-Roda (1964b, 1969c) es- timated that 500,000 age bluefin were taken by the traps at Ceuta and Mediterranean Morocco and the Ceuta fishing fleet in September-Oc- tober 1963, but stated that such ex- tremely numerous catches of these fish were exceptional. Crespo and Rev (1976) showed that the catches of age bluefin were much greater numencally than all other sizes com- bined in eight of the 14 seasons from 1961-1974, with significant highs in 1963, 1967 and 1973 Rey el al. (1977) reported that 22 bluefin were caught in the Ceuta trap in 1975 Twenty of these were reportedly more than 4 years old, and weighed 12,50(1 kg. Perhaps there is an enor in these figures, as the average weight would be 625 kg. We have little information on the bluefin tuna trap fisheries of Al- geria. Heldt (1932a) listed six traps set in 1 930 and two in 1 93 1 , but with little success in most instances. Belloc (1961) named three in the vicinity of Oran, and listed their individual yearly catches for 1953-1958. The largest such catch was 69 tons (all species) and the average was about 30 tons (all species). Specialized nets used off the French coast, and their catches, have been described by du llamel de Moneeau (1769-1782), Doumenge (1953), di Meglio (1962) and Farrugio ( 1 977). These included fixed and drifting gill nets and beach and pelagic multiboat seines, called n NUMBER OF FISH/10^ ■ METRIC TONS/IO' r20 -15 -10 Year Figure 51. Annual catches of norlheaslem Alhuilic fisheries I'oi large and medium size bluefin tuna (Scandinavian and Noilli Sea seine and liiHik-and-line fisheries and Ibero-Moroccan Hay trap lisheiy) in lens of thousands of li.sh and lu thousands ol' ions "thonnancs" or "thonaires," "eombriercs," "eourantilles." and "scinches " In some cases the same name was applied to \arious gears, according to the locality. The name "thonnaire" appears to have been used for several types of gear Hook and line gears used off the French coast were described or men- tioned by du Hamel de Moneeau (1769-1782), Doumenge (1953), di Meglio (1962) and Farrugio (1977), These included longlines, trolling gear, poles, and sport gear. Dieuzeide (1931) described handline fishing for large and medium bluefin in the Bay of Castiglione, Algeria Thomazi (, I 947) and Dicuzeide (1 949) detailed an interesting method of hook and line tuna fishing utilizing small fishes, which had gathered around a moored branch of a tree, as live baits. They reported that this method was used off the Mediterranean coasts of Spain and Morocco, as well as off Algeria Recreational fishing for bluefin tuna has been de\eloped off the Medi- terranean coasts of France and Spain. Commercial fishermen have been catching medium and giant bluefin by trolling with rod and reel gear off Port de Bouc, at the mouth of the Rhone, since 1959. The season ex- tends from late June into early No- vember, Annual contests, in which sport as well as commercial fisher- men participate, have been held. In one which took place August 27-31, 1 969, 1 5 boats landed 40 tuna with an a\eragc weight of 105 kg, even though the weather permitted fishing on only two days. The largest bluefin taken weighed 220 kg. The above information is from Gianelli (1969) Another French port where recrea- tional fishing is developing is La Grand Motte, on the Gulf of Lions, where tuna up to 30-40 kg are taken (Cesarei) 1972). Sport fislimgofi'the French Riviera will be discussed in the folkwing subsection Recreational fishing has also be- come popular on the Spanish Mediterranean coast at Rosas, near the French frontier, and at Castellon Catches at Rosas include small tuna and giants of 150-180 kg (Cesareo 1972) Tuna tournaments are held annuallv at Castellon in late August and earh September Catches totaled 50 50° ACT 30" ©CEOCRAPHICAl. MCFCRENCCS t.CIU*»W LMICIMTl I.UICm.UMI «.IM.flUIICI«. •.eCNOA TCOIflC* 10. ii.fiiimE u. isiMtam. ncam M. ova us I9.CUM or nf mm H.iuini If.nMPOLI M.TUM( llWCICDS 12.CU.F or c*snciK)N( u.cuir or *b»u M.OMM 20° 10' Figure 52. Geographic references for the Mediterranean and Black Seas. Figure 53. Geographic references for the northwestern Mediterranean Sea. 47 fish weighing 4,536 kg with maxi- mum, average and minimum weights of 161, 96.5 and 38.2 kg, respec- tively, in 1 973, and 32 fish with maxi- mum, and minimum weights of 180 and 37.8 kg, respectively, in 1974 (M. R Borrcll and \l. K Harry, per- sonal communications). In 1976, however, only one fish was caught (I. F dc Gamboa, personal com- munication) Tuna of the abo\c sizes were taken ni August and early Sep- tember, but after they disappeared, great quantities of age bluefin (about 1 kg) were taken from late September to early November (J. Hcxnat Orti/, personal communica- tion) 51 The authorization of seining for blucfin tiinn in iTcnch Mediterranean waters in 1960 brought about a eon- siderablc increase in the catches of the species The development of the fishery in the Gulf of Lions was de- scribed by di Meglio (1962) and Farrugio (1977). Some old boats, mainly trawlers, were fitted with this gear The first seines, "cinciole", were derived from the Italian type and the second, "cerco", from the Portuguese type used off Morocco The annual catch in the Gulf of Lions immedi- ately increased from about 500 tons in the years 1 950- 1 96 1 to about 1 ,000 tons in 1 96 1 and 1 962 (Maurin 1 964) Most of the 1 962 catch was taken in the beginning of the year in the re- gion of Marseille and Rousillon The French seine fishery for blucfin off Pro\ence and in the Gulf of Genoa began in 1967 (Patania 1967) and has produced more than half of the total French Mediterranean catches of this species since then (Farrugio 1977). Hamre et al. (1971) reported that from July 1969 through Januan,' 1970 purse seine catches off the Mediterranean coast of France to- talled 1,500 tons, with most catches occumng in October and No\'ember. Catches have fluctuated around 1 ,500 tons since 1966 with lows of 1,100 tons in 1970 and 1972 and a high of 2,200 tons in 1971 (Farrugio 1977). Miyake et al. (1976) reported gener- ally smaller catches in this period For some years, their figures differed considerably from those of Farrugio (1977). The best pre-seining fishing in the Gulf of Lions was from earl}' April to mid- June and from late July to the end of September, with a lull from mid-June to mid-July. Scattered fish were caught in hitc Februai-y and March, and from October into early December. Late December and Janu- ary were periods of scarcity (Doumenge 1953) Data on the 1963- 1965 catches (Mamie et al 1968) showed minor peaks in March and major peaks in Scplembcr-October Farrugio (1977) reported that the seine fisheiy was active in the Gulf of Lions from early March to late May and from October into Decem- ber and in the eastern area (St Trope/ to Capo delle Melc) from .luly to late September or early October Farrugio (1977) described the seine fishing areas. In the Gulf of Lions, fish were taken in September and October in two areas, one be- tween Port Vendrcs and Leucate and the other between Sete and Marseille Beginning in November or Decem- ber, they tend to concentrate in the northern part of the Gulf between Sete and the mouth of the Rhone At all seasons, the fish are generally less than 20 miles (33.3 km) from the coasts, in waters whose depths rarely exceed 100 m OlT Provence and in the Gulf of Genoa, the fishing occurs from 5 to 50 miles (8.3 to 83 km) off the coast between Cap de S. I'ropez and Capo dellc Melc Until recently, little information on the sizes of fi.sh taken in these fisheries has been available. Gounet ( 1 894) showed that the average sizes of blucfin taken near Marseille in 1891 and 1892 by traps and other gears ranged from 18 to 28 kg. Samples of catches taken in the Gulf of Lions (off Languedoc, l-rancc) in 1953 and 1954 consisted of blucfin of ages 1, 2, 3, and 4 (about 48, 80, 97 and 119 cm long, respectively) (Doumenge and Lahaye 1958) Little variation in sizes offish with method of capture was indicated by cither of these studies. Samples of the 1968 and 1969 catches showed that most of the fish were in the "small" size group, with age 3 dominant in both years (Figure 54). Since then, this predominance of age 3 fish has re- mained relatively constant (Figure 55) In 1970 and 1971 a significant showing of medium (ages 6-8) and small giant (ages 8-10+) blucfin oc- curred but this has not happened since (Farrugio 1977) The 1971 catch con- sisted partially of 120 tons of blucfin weighing from 100 to 150 kg (ages 8-10), which were reportedly taken o(\' Sete, France, in April by three sardine-tuna seiners (Anonymous 1971). fhc possibilities of seining blucfin off the Meditenanean coast of Spain were mentioned by Hcllon ( 1954), but, apparently, they have not been developed significantly In their review of the fisheries for tunas and related s|iecies m the Mediterranean and Atlantic waters off southern Spain, Laboratono Oceanografico I'alma cl al (1976) do not mention seining, c\'en among the minor gears. Rodrigucz-Roda (1964a), however, noted that considerable numbers of age blucfin (0.5 kg in September, October, 1 kg in November, and 2 kg in December) were often taken by sardine .seiners off Alicante, Spain, mixed in with the sardines Consid- erable catches of age blucfin in the same season during 1976 were also reported (F L de Gamboa and J Hexnat OrUz, personal communica- tions). Rodriguez-Roda (1964b) also reported considerable catches of age blucfin m the fall of 1963 off the Mediterranean coast of Morocco by Spanish seiners based at Ceuta. Spanish and Japanese longlinc \essels ha\e also taken bluefin tuna in the western Mediterranean An- nual catches of the Spanish Mediter- ranean longline ficet, which fishes primarily for broadbill swordfish, Xipluas gladius Linnaeus (1758), in- cluded from 16 to 274 tons of bluefin in the years 1968-1974 (Laboratorio Oceanografico Palma et al 1976) Japanese longhners took 238, 427 and 7,980 bluefin tuna in Mediterra- nean waters west of longitude 10°E in 1972, 1973, and 1974, respectively. Fish were caught from April into October, with the highest catch rates, up to 24 fish per 1 ,000 hooks, occur- ring in June. Good catch rates were also attained in May and August (Fisheries Agency of Japan 1974. 1975, 1976). Rodriguez-Roda (1969d) ex- pressed concern over massive catches t)f very N'oung blucfin in the western Mediterranean, as w ell as off the At- lantic coast of Morocco, which he had described previously (Rodrigucz- Roda 1964a, 1964b) He reported catches of large numbers of age fi.sh in the fall by seiners off Alicante and the trap at La Linea in Spain, and by .seiners and Irajis off the Mediter- ranean coast of Africa ]ust east of Gibraltar He reported that, in the period September 15-October 19, 1963, about 500.000 age bluefin were taken by the Ceuta seiners and the traps in the \icinity This was. however, an exceptional year, accord- ing to his infomiant. Wc have been advised that comparable catches of 52 age bluefin were taken off Castellon, Spain, in the fall of 1976, mainly by sardine seiners, but also by sport fishing boats. Rodrlgucz- Roda (1969d) believed that the de- cline in the Spanish Atlantic trap fish- eries for larger bluefin might have been caused by excessive fishing of immature fish. Bluefin tuna catches in the west- ern Mediterranean were evidently un- important prior to the development of the French seine fishery' in 1960 Combining the data of Miyake et al. (1976) and the Fisheries Agency of Japan (1974, 1975, 1976), it appears that the total western Mediterranean catches varied between 1 ,000 and 2,500 tons (approximately) in the years 1965-1973, but rose to about 4,200 tons in 1974 The latter figure was due mainly to exceptionally large catches by the French purse seine and Japanese longline fisheries. The fishing of age bluefin in this area, which apparently is very poorly re- corded, appears to be a cause for concern. b. Central Mediterranean The central Mediterranean (Fig- ure 56) has traditionally been the area of the major bluefin tuna catches in that Sea. The tuna trap has been the dominant gear in the area until recently The oldest trap fisheries were around Sicily and Sardinia, and off the southwestern coast of Italy itself Additional traps were subse- quently installed off Tunisia and western Libya. All of these traps were originally of the Sicilian type (Fodcra 1964). Changing conditions have ne- cessitated their modernization through the use of better materials and the adoption of the Spanish de- sign. This process began in Tunisia in 1950 (Anonymous 1952) and in Sicily in 1956 (de Cristofaro 1970). In addition to the ancient trap fisher- ies (Pavesi 1889,Parona 1919,Belloc 1961, Sara 1964), bluefin tuna are taken in the central Mediterranean by purse seine (Scaccini and Biancalana 1959, Paini 1975), hook and line (Scordia 1931, 1932; Genovese 1965, Cesareo 1967) har- poon (Scordia 1932, Sara 1968) and longline (Sara and Arena 1967, Shingu et al. 1975). In the early kj 10 < j>^ o '^ "^ X X '^ — VC "^ "^ <; o o —T <~^ ^ "P'-B'-B'-B >N— _>, ^ m 5 -c^ :5 Z 1/3 1/3 !/J M W > ■a no 3 C d u ^ ^ v=^ "^ '£. § So o a o sJ -a o c o .a J (Z) a. a. « cu 2 .ta < o o o X o o X X XXO xxs X X XXg; xxg XX5 XXg; On X X m ON 0\ On CN| W-. O NO r- — I S3 .a in CO ^5 3 'm o § g .a o > e 1 •^ O s tu ti- l« u •o k. J. X o •a &u as c o c O C' c o o c g XXX NO ' g XXX ' X X X X X '2? XXX XXX ' » X X X ° ' pi X X X X ■ s; X g X X r- — — On NC »/^ On On C O O O C-l — — ' o o o ON m oo o o o r*^ oo .a .a -s -a ^ 00 on 17 .^ .a CO ^ ^ « izi in £ g > W BJ tJ Z Z c« t/i M « ra — 2 -r; ^ 0. ^ C/3 C« O C/3 U -a C -a fS s o o o o O 00 re '2 S ^ '"3 C 1^ C u u •— -J C § -I 1 « c -I s yl ^ O is .^ "O c« « C U '-« CO § 1 J — , =^ 3 25 ffl u i ° ■/! I, t II o ? T3 "6 E U (X u- o S ? 5 on o ,2 -e o ■u CU 00 re re > DO re a 'A 0. o ' Arena confirm the indications, also noted in the average weights of the trap catches, that there was a substantial recruitment of younger fish to the central Mediterranean spawning stock in 1975. The size composition data suggests a strong influx offish in the 40 to 1 00 kg class in that year. This is most encouraging, but whether the stock can withstand the massive im- pact of the new seine fishery remains to be seen. As noted above, Paini (1975) re- ported that the large seiners, after completing their fishing in the Tyrrhenian Sea and the Sicilian Chan- nel, moved north to fish small tuna and aibacore. P. Arena (personal communication ) advised us that sein- ers had taken about 2.600 tons of bluefm weighing from 10 to 100 kg in the northern Tyrrhenian and Ligurian Seas in September and Oc- tober 1976. Thus very heavy pres- sure is also being put on the stock of young fish just as it appeared to be furnishing urgently needed recruit- ment to the spawning stock of large bluefin. Although Ninni (1921a) and Scordia (1939) had reported occur- rences of bluefin tuna off the Italian Adriatic coast near Venice and near Manfredonia and Molfetta, respec- tively, it was not until 1950 that an Italian Adriatic tuna fishery was ini- tiated (Scaccini and Biancaiana 1959). This purse seine fishery has been described by Levi (1977). The season extended from the end of March to mid-November, and the tlshing area covered most of the coast between Barletta and Trieste (Fig- ure 58). The most successful fishing occured in March-April off Pescara and Punta Penna, in the central Adriatic, for fish between 10 and 50 kg and in August-October, to the north, off Porto Garibaldi and Cattolica-Cesenatico for 6 to 10 kg fish. The fishing originally took place about 15 miles (24 km) or less from the coast, but this distance has gradu- ally increased to 30-40 miles (48-64 km). The yearly catches have varied between 83 and 434 tons, with an average of 186 tons, from 1955 through 1971. Levi (1977) reported that two of the larger and more mod- ern Italian seiners had entered the Adriatic fishery in 1976. P. Arena (personal communication) advised us that the 1976 Italian seine catch in the Adriatic had reached about 1 ,000 tons in October Hook and line fisheries have ex- isted in several parts of the central Mediterranean, but the most impor- tant was in the Strait of Messina. Scordia(1931, 1932, 1934, 1935)and Genovese ( 1 965) have described this fishery, which embraced two distinct seasons. The fall-winter fishery (Sep- tember-March) occurred in the north- em part of the Strait of Messina, and took small or medium sized fish be- tween spawning seasons. TTie sum- mer (late June- August) fishery oc- curred in the southern part of the Strait, and took giant fish which were maturing or recently spent. Scordia believed that the winter fish came from the Tyrrhenian Sea, and the summer ones from the Ionian. She provided extensive data on the catches for 1 928- 1 935. Genovese pre- sented the size composition, by 59 30 20 - <0 30 20 10 ««j 30 S 20 30 20 10 30, 20 10 30 20 10 1958 1965 -£7- 1966 '-f.?- 1967 1968 1968 1969 1969 ■£2- 1970 1970 1971 i-^^ 100 200 LENGTH (cm) 300 150 200 L ENG7H (cm ) Figure 57. Lengths of bluefin tuna t;iken in the Sicilian trap catches in 1958 ;md 1965-1971 ("+" on graph means less than 0.5%). weight groups and months, of the catches for most of 1948-1950 and 1953-1954, with the sex ratios for some samples. The bluefin taken in the winter fishery were mostly in the medium size group, but with many small t'lsh and a few giants. The in- troduction of live-bait chumming in 1950 increased the catch from a few hundred fish per year to as many as 1,500 in a montlr, but after 1953 the catch declined. The summer catches were nearly all giant fish, with a few mediums. Only from 23 to 44 fish per year were taken in 1948-1950, and none were caught in 1953-1954. The sex ratios of the tuna taken in these years showed a veiT slight predominance of females (about 52% females), which did not van,' greatly between months and years. Arena (1959a) described a productive line fisher)' off tiie Aeolian Islands (north of eastern Sicily). Up to 27 fish were taken per boat day under favorable conditions. It was necessary' to use fine lines baited with sand launces, Ammodytes cicirella. The fish taken generally ranged from 30 to 40 kg. Much larger individuals, 200 kg and over, were often hooked, but usually broke the light lines. The Italian longliners fished primarily for broadbill swordfish, but caught blue- fin tima incidentally (Sara and Arena 1967). Their tuna catches have not been recorded in (he statistics. R. Sara (personal conimunicalion) estimated that the swordfish longliners might have captured "not over 500" tima with an average weight of about 1 50 kg during the 1972 season. Sara and Arena (1967) noted that smaller longlines used primarily for dolphin (fish), Coiyphaena hippurus Linnaeus (1758), also took unrecorded numbers of small blue- fin. 60 The Japanese longline fishery en- tered the Mediterranean in 1 972, tak- ing 459 bluefin between IO°E longi- tude and 20°E longitude. Its catches in this area increased to 748 fish in 1973 and 4,914 in 1974 (Fisheries Agency of Japan 1973, 1974, 1975). Assuming that the weight of these fish was the same as the average for the entire Mediterranean, the cor- responding tonnages would be about 74 in 1972, 156 in 1973, and 835 in 1974. The numbers offish taken in 1975 are not available, but the total Japanese Mediterranean catch de- clined from 2,192 tons in 1974 to 1,100 tons in 1975, because of con- servation measures (Kume 1976). The largest catches were taken south of latitude 40°N in May and June, corresponding to the "arrival" run of maturing bluefin in the area. Catch rates, in fish per thousand hooks, var- ied from to 12.8 in May, 7.8 to 14.6 in June, and to 19.5 in July. Thus the most consistent catch rates were obtained in June, during the arrival run when the maturing bluefin are believed to be reluctant to feed (see Section VI). The highest catch rates, 14.6 and 19.5 fish per thousand hooks, were obtained in the Ionian Sea in June and July, respectively. Sport fishing for small bluefin tuna and albacore ( Thunmis alalimga) in the Ligurian Sea was initiated by French and Italian anglers in 1962 and 1963. This fishery was very suc- cessful in the years 1964-1968, with daily catches in 1964 averaging 30 timas a day for a well crewed boat with expert anglers. This average de- clined to in 1971, reportedly be- cause of heavy fishing by Sicilian and French seiners in 1968-1971. Some improvement in the fish- ing was noted in 1972 and 1973, af- ter the seiners had left the area. The schools were much less numerous than before the seining, however, and were present for a much smaller por- tion of the year (Cesareo 1973,"A.C." personal communication). This im- provement has evidently continued; nine boats took 680 kg of bluefin and albacore in a day and a half in a tournament off San Remo, Italy, in September 1975 (di Sant' Ignazio 1975). Friction between seiners and recreational fishermen was also men- Table 15. Catch data for the Italian purse seine fishery in the southeastern Tyrrhenian Sea and the Sicilian Channel (Miyake 1976, P. Arena, personal communication). Year Boats Fish Tons Average Weight (kg) 1972 8 9,000 2,300 256 1973 11 8,500 2,200 259 1974 15 13,000 3,500 269 1975 16 32,000 5,800 181 1976 4,120 tioned in this report. Cesareo ( 1 974b) stated that there had always been gi- ant bluefin along the Italian Adriatic coast from the delta of the Po south- ward, from Punta Pila to Porto Corsini (Figure 58). They were especially numerous in 1968-1970. Catches were scarce, however, because of the lack of sport fishing effort. Cesareo's comments were in response to a let- ter from a sportsman who reported the capture of a giant fish weighing 176 kg off Punta Pila in 1974, and recalled taking one weighing 1 52 kg in the same area in 1971. Scaccini (1961) discussed the distribution of young bluefin tuna in the Adriatic in relation to physical, chemical and dynamic conditions of the environment (Figure 58). Very small fish (11-20 cm. 40-100 g) oc- curred in summer close to the coast from north of Rimini to Ancona over sandy bottom. Fish 40-60 cm long (3-5 kg) and larger ones 10-12 kg and up to 1 m long were fished by seine from April to September in the north Adriatic from the mouths of the Po to just north of Ancona, al- ways from 6-7 to 12 miles offshore. On the other hand fish of the same size were found from one half to five miles off the coast in very shallow water in the middle Adriatic between San Benedetto de Tronto and the Gargano promontory. They also oc- curred in the mid-Adriatic, always on the surface, north of the island of Pianosa over the trench of Porno. Bluefin of 30-70 kg and more than a meter long, rare in the western Adriatic, were caught habitually in the eastern Adriatic, and in summer and early fall north of the mouths of the Po near Venice. Scaccini showed that this apparently strange pattern was based on currents whose waters were more favorable than the sur- rounding ones for the tuna of the sizes in question. The oldest bluefin tuna fishery in the Adriatic is by traps situated along the northern coasts of Yugo- slavia and the adjacent islands. As elsewhere in the Mediterranean, the number of traps fishing off Yugosla- via has decreased greatly. Whereas Parona (1919) listed 38 active em- placements, only 21 existed in 1957 and 17 in 1958-1959 (Belloc 1961). Table 16. Size data for catches of the Italian purse seine fishery inthesoutheastem Tyrrhenian Sea (P. Arena, personal communication). Percentage of Percentage of Size Range (kg) 1974 Catch 1975 Catch 40-100 kg 30% 100-200 kg 20% 25% 200-300 kg 50% 35% 300-500 kg 30% 10% Figure 58. Geographic references and movements of bluefin tuna along the Italian Adriatic coast. A Yugoslavian purse seine fishery for bluefin tuna was introduced in 1929 (Tilic 1954), with seven ves- sels fishing in the period 1936-1940 and 1 1 in the period 1 947- 1 95 1 . Like the trap fishery, this occurred mainly off the northern coasts and the adja- cent islands (Morovic 1961). Total yearly bluefin tuna catches in Yugo- slavia averaged 127 tons for 1936- 1940 but increased to 468-873 tons in 1947-1951 (Tilic 1954). In 1952- 1974, the total yearly catches varied from less than 100 to 700 tons, but have not exceeded 350 tons since 1959 (Hamre et al. 1966, Miyake et al. 1976). Trap fisheries have long been important in Tunisia (Figure 56). Al- though as many as 1 1 traps have op- erated in some years (Gaudiiliere 1954), Roule (1924) recommended that studies of the bluefin tuna stocks in the area be concentrated on the catches of the Sidi Daoud trap on the west side of the Cape Bon peninsula. This trap has been by far the most productive in Tunisia, and has fished almost continuously since 1 863 (Fig- ure 59). The other traps have fished intermittently and their catches seemed to follow the same trends as those of Sidi Daoud. Its catches for 1863-1923 (excepting 1874 when it was not set) averaged about 8,000 bluefin tuna per year (Roule 1924). In the years 1928-1937, regarded as a period of crisis, its annual catches ranged from 1,000 to 3,400 fish (Heldt 1932, 1934, 1937. 1938). In 1955 it took 3,600 fish with an aver- age weight of 128 kg (Postel 1956). In the period from 1962 through 1976, its approximate average an- nual catches declined from 1 ,500 fish weighing 180 tons in the first five years to about 300 fish weighing 40 tons in the last tlve (M. /aouali, per- sonal communication), fhus, the Tu- nisian trap fishery for bluelln has apparently suffered a collapse simi- lar to that of the Italian trap fisheries. Sella (1929a) pointed out that the trends of the Sidi Daoud catches were remarkably similar to those of the important traps off southwestern Sardinia. Roule ( 1 924) concluded that the Sidi Daoud catches varied in- versely with the amount of rainfall in the area. He felt that the runoff from the Lake of Bizerte reduced the sa- linity of the waters around the trap, making the area less attractive to the bluefin. Total catches of the Tunisian traps in 1 9 1 0- 1 923 ranged from 4,300 to 34,400 fish with the larger catches being taken before 1916 (Roule 1924). In 1927-1938 catches ranged from 2,200 to 9,000 fish weighing 200 to 800 tons with yearly average weights of 7 1 to 1 08 kg (Heldt 1 932, 1934, 1937, 1938). The three traps operating in 1955 caught 3,985 blue- fin averaging 1 19 kg (Postel 1956). This constituted only 47 percent of the total catch of the traps, by weight. The annual total catches of all spe- cies by the three Tunisian madragues active in 1 952- 1 958 ranged from 677 to 1,013 tons(Belloc 1961). Postel's ( 1 956) data suggest that about half of these tonnages might have been blue- fin. Total Tunisian bluefin catches for 1964-1974 ranged from 200 to 900 tons with smaller catches occur- ring since 1970 (Miyake et al. 1976) but no breakdown by gear is avail- able. Heldt (1932. 1934, 1937) re- ported that catches of up to 40 tons a year were taken in the "winter fish- ery", mainly with seines. The Tunisian traps all fish in the "arrival" run, except for some ex- perimental fishing of the "return" run by the Cap Zebib trap in 1 924 (Gruvel 1926) and an experimental trap at Ras Mustapha (Bellon 1954). For 1931-37, the first trap catches were from Ma\ 19 to June 9, and the last from .lune 20 to July 6 (Heldt 1932, 1934. 1937, 1938). Heldt(1929)sup- plied average weights of the bluefin caught by the Sidi Daoud and Ras el Ahmar traps and tor all the Tunisian traps for 1904-22. The average for all traps ranged from 50 to 89 kg. The average for Ras el Ahmar (59- 126 kg) was higher than that for Sidi Daoud (50-89 kg) in every year but two. Since Ras el Ahmar is situated north of Sidi Daoud, nearer the tip of the Cape Bon peninsula, Heldt con- cluded that the larger bluefin trav- elled farther offshore than the smaller individuals. Heldfs (1932, 1934, 62 1937, 1938) data for individual Tu- nisian traps showed that those on the north coast (Cap Zebib, Sidi Daoud, Ras el Ahmar and El Aouaria) caught larger fish than those on the east coast (Monastir, Kuriat, Conigliera, and Bordj Kadidja). Yearly average weights for the western traps varied from 42 to 246 kg, whereas those for the east coast traps ranged from 47 to 62 kg. Posters (1956) data for 1955 catches showed the same trend: 125 and 115 kg for Sidi Daoud and El Aouaria, respectively, and 52 kg for Kuriat. Sella (1912a) reported that there were quantities of bluefin tuna along the Lib_\an coast, and the installation of traps began in 1914 (Parona 1919, Ninni 1921b). Data are available for the catches in Tripolitania (western Libya) in 1915, 1919-1936, 1939, 1951, 1955, 1959, 1972, and 1973, but not all on a comparable basis. The most complete data are for 1 92 1 - 34 (Anonymous 1928a, 1929a, 1932a, 1934a, 1934b, 1935a, 1935b, 1936). In 1915 and 1919-20 from one to three traps were set each year, and annual catches ranged from 1 , 1 60 to 6,206 fish per trap. In 1921-34, from six to 13 traps were set each year, and annual catches varied from 285 to 1 ,200 tons per year, consist- ing of from 547 to 2,067 fish per trap with average weights of from 56 to 1 10 kg per fish. In 1935, 1936 and 1939 from seven to 10 traps fished each year and took from 1,133 to 1 ,737 fish per trap year. The average weight of the fish caught in 1 936 was 75 kg. Catches totalled 1 , 1 20 tons in 1951 (Anonymous 1952), 6,403 fish averaging 86 kg in 1955 (Postel 1962), and 1,950 tons in 1959 (Anonymous 1960). Recent landings have been considerably smaller, with four traps taking 1,550 fish in 1972, and three taking 2,360 in 1973 (R. Sara, personal communication). The total yearly bluefin tuna catches of Libya for 1964-1974 varied between 300 and 2,000 tons (Miyake et al. 1976), with the largest catches oc- curring in 1968 and 1969. The Tripolitanian traps fished the "arrival" run only, although evidence of a potential "return" fishery has been noted (Anonymous 1932). First catches for 1927, 1928, and 1931 occurred between May 29 and June 23 (Anonymous !928a^ 1929a, 1932). Belloc (1961) reported that the fish- ing began at the end of May, peaked in the first half of July and ended during the last half of July. Sara (1964) stated that the Libyan traps fished about 15 days later than those of Sicily, Sardinia and Tunisia. In 1964, the fishery extended from the 22nd to the 26th week (about May 25-June 23) (Hamre et al. 1966). The only size sample available for the Libyan fishery is for the 1964 catches (Figure 60). About three- quarters of the fish were in the "me- dium" category, with the remainder in the "large" group. Except for an ephemeral attempt near Tobruk (Sella 1932a), the east- ernmost tuna trap which fished spawning bluefin in the Mediterra- nean was at EI Mongar near Bengasi in Cyrenaica, Libya. Information on this trap is meager. In 1924-27, it took from 1 ,039 to 3,286 bluefin tuna per year (Anonymous 1928b). The latter figure, attained in 1927, was reportedly the highest for any trap in the Mediterranean in that year. In 1928, only 197 tuna were taken (Anonymous 1929b). Catches of from 1 12 to 436 tuna per year are listed for Bengasi or Cyrenaica for 1930-33 (Anonymous 1932a, 1934a). In 1 927, the first and last bluefin catches at El Mongar were on May 31 and July 18, respectively (Anony- mous 1928b), and in 1928, on June 26 and July 14 (Anonymous 1929b). The latter season was regarded as unusually late in starting. No quantitative data arc avail- able on the sizes of the fish taken, but Sella (1932a) stated that they were small or medium size, and that their gonads were ripe. Anonymous ( 1 929b) reported that the tuna caught in 1928 were very small. Because of the irregularity of its catches and the small size of the fish taken, Sella (1932a) did not regard El Mongar as a true "arrival" trap. Since it did not catch large spent fish, he also felt that it did not qualify as a "return" trap. It appears to have been abandoned soon after this information was published. He pointed out that this trap was east of the 38 o/oo isohaline, which he regarded as the upper limit of salinity for large "ar- rival" (maturing) bluefin. He added that, had he known what he then did about the sensitivity of maturing blue- fin to salinity, he would not have attempted to establish a trap east of El Mongar (in water of higher salin- ity) near Tobruk. c. Eastern Mediterranean and Black Sea Bluefin tuna are extensively dis- tributed over the eastern Mediterra- nean (Figure 52), except its south- east corner, but the only fisheries of any importance are those of Greece and Turkey. Annual catches reported for Greece since 1952 have ranged from less than a ton to 1,220 tons, and those for Turkey since 1 957 from less than a ton to over 1,500 tons. Catches in Greece since 1968 and in Turkey since 1970 have been negli- gible (Hamre etal. 1966. Miyake and Manning 1975. Miyake and Tibbo 1972). Eleven traps were active in Greece in the period 1954-1958 (Belloc 1961). The maximum annual catch (all species) per trap was 21 tons; few catches exceeded 10 tons and many were less than one ton. All of the traps were located on the Aegean (eastern) coast of Greece (Figure 61). with al I but two on the Peleponnesus. Ninni (1922) quoted Vinciguerra (1 896) to the effect that the only bluefin tuna fisheries in the Aegean were at Melina in the Gulf of Volos and Gialtra on the island of Euboea. Ninni concluded that in spring bluefin tuna migrated north- ward in two groups through the Aegean from a wintering area around Crete. The major group skirted the coasts of Asia Minor and the adja- cent islands and a smaller one tra- versed the channel between Euboea and the mainland to enter the Gulf of Volos. Athanassopoulos(l923, 1924, 1926) considered the bluefin to be a rare fish in Grecian waters, particu- larly off its western (Ionian Sea) coast. Oren et al. (1959) reported trolling catches of very small (45-53 cm total length) bluefin tuna in sev- eral parts of the Aegean during an exploratory fishing cruise Septem- ber-December 1952. 63 «n r15 10- -10 5- 0-" .-r- 1860 65 70 75 80 85 90 95 L 1900 19n 10 5- 4WhJU 30 35 40 r15 -10 -5 1901 05 10 5- 41 45 50 60 65 D NUMBER OF FISH/10^ METRIC TONS/IO^ 60 Year *4tt 70 75 T — r- 60 r 10 Figure 59. Annual catcli of bluefui tuna tVoin llic Sidi Daoud trap in I'unisia in tliousands ol lisli mid hundreds ot" metric tons. Turkish fishing for bhietln tuna is by trap, handlinc and harpoon (Ninni 1922. Dcvcdjian 1926, Hovas.se 1 929. Lebederfl9.-^r,.Akyuz and Artuz 1957. lyigiingiir 19,^7. Belloc 1961). Lebe'deff (19.16) de- scribed sport fishing near Istanbul in the winter of 19.15-.16. The favorable situation for developing a sport llsh- ery there was also mentioned by lyigilngor (1957), but we have no further information on this subject. I'arona (1919) showed 26 loca- tions for traps in the easlem .Sea ol' Mannoia and the Bosphorus. 13elloc (1961) noted that Devedjian (1926) showed 225 emplacements lor 1 urk- ish traps in the Black Sea. Bosphorus. Marmora and Dardanelles, but his charts show only I 9 in the Bosphorus and eastern Marmora and 1\\() m the Dardiuielles. Only a few of these were noted for catching bluelln tuna. Ninni (1922) noted that there were .some "summer" and .some winter' traps. but lyigiingor (1957) said that the traps fished from early April to the end of August. He estimated the av- erage season's catch per trap at 100- 150 mna. Opinions al.so dlifered as to the periods of pa.ssage. Ninni de- scribed an "arrival " (northward) run from late March to mid-August and a "rctuni (soullnvard run from early Augiisl lo December or even I''ebru- ary. llo\asse (1927). on the basis of catch records for the years 1915 and 1921-192''. Ibund three maxima in 64 30 20 10 "-ZZr 50 100 150 200 250 LENGTH (cm) Figure 60. Length composition of samples of bluefin tuna taken along the Libyan coast in 1964 (" than 0.5%). ' on graph means less the landings: one with maturing fish in February, March and April, one with spent fish in July and August, and a lesser one in December. Hovasse believed that the first run was an "arrival" migration into the Black Sea, where spawning took place and the second was a "return" run fi-om the Black Sea. He did not understand the significance of the third maximum. lyigUngor (1957) listed three rather similar runs: March-April, July-autumn, and No- vember-January. Akyuz and Artiiz (1957) specify an "arrival" run simi- lar to Ninni's, from April to early August, and a "return" run from late October into December. These au- thors show the average volume of bluefin tuna sales in the Istanbul mar- ket, by month, for 1935-1955. These show a maximum in March and an- other in July, with good volume in April, February, and August and a minimum in June. The yearly vol- ume of bluefin tuna sales in the Istanbul market varied from 85 to 537 tons in 1909-1916 (Ninni 1922), from 15 to 112 tons in 1928-1938 and from 40 to 764 tons in 1953- 1955 (Akyuz and ArtUz 1957) The allocation of catch by gears is not shown. Harpooning is practiced both from the shore and from small boats (Devedjian 1926). Handlining is also practiced from small boats (Devedjian 1926, lyigUngor 1957). The hook-and-line fishing seasons are November-Januar>- and March-April; the fish do not take bait July-October (lyigUngor . 1 957). Both of these fish- eries are centered around Istanbul. Ninni (1922) concluded that the quantity of bluefin tuna available was much greater than was indicated by the catches of the traps, since the latter did not extend into sufficiently deep water to fish tuna effectively, and also were too weak structurally to hold any considerable quantity of bluefin. He noted that a trial setting of an Italian style trap at Touz Bumum in 1913 had not produced satisfactory results, but did not con- sider this experiment conclusive. Evidently most of the bluefin taken near Istanbul are large. Most of the fish examined by AkyUz and Artuz (1957) at the Istanbul market (1955-I956)wereinthe 170-280 cm range, with the smallest 120 cm and the largest 330 cm. Samples exam- ined there in 1967 and 1968 (Hamre et al. 1971) ranged from 85 to 320 cm, with most of the fish from 155- 290 cm (Figure 62). Lebedeffl 1 936) mentions that hook and line catches in 1 936 included 40 bluefin over 400 kg, 10 over 600 kg, and some from 700 to 775 kg. He noted that fish over 200 kg were taken only in the fast current, whereas 100 yards away only individuals of 20-50 kg were caught. In early April, the large and small bluefin left, but medium sized ones stayed near Prince's Island. Dur- ing a few days of sport fishing, from December 25, 1935 to late March 1936, Lebedeff caught tuna of 120, 230, 350, 380 and 400 kg. He also lost several large bluefin which broke his lines or leaders. One of these, which was later caught in a net and identified by the hook and leader which had remained attached to it, weighed 450 kg. Hovasse (1927) noted that blue- tin tuna occurred in the Aegean Sea- Marmara-Black Sea area in waters of extrenicK different salinities, from 1 8.3 o/oo in the Black Sea to 39 o/oo in the southern Aegean. In the vicin- ity of Istanbul tuna are found simul- taneously in deep waters with salin- ity of 38 o/oo and shallow waters with only 17 o/oo to 18 o/oo. He stated that tuna in the area were found in considerable ranges of tempera- ture—as low as 14°C to 19°C, ac- cording to depth — in the Black Sea at the time of the "arrival." These findings conflicted with Roule's (1924) hypothesis of a stenothemiic and stenohaline tuna. Small bluefin tuna range along the Mediterranean coasts of Turkey and Syria, and occasionally occur off Cyprus (Carp 195 1. Oren et al. 1959, Sara 1904). Carp reported a Turkish fisher) for bluefin (apparently 10-30 kg) along the Anatolian coast and in the Gulf of Mersin. Oren et al. ( 1 959) reported captures of very small (45- 65 AciTICS 0SCOait*PMCU. KCrcRCNCCS I.PtLOPOMNEtUt •.OtHOMEUES Z.ikTHEMI '"AS'.. J.CWOIA — •— •» 4.VMLrl>* s.euir OF votos •.PKINCCS IS. lO.BOSPOIIUS ly ai" Figure 61. Geographic references in the Aegean Sea area. 30 r 20 5 it 30 20 10 O 1967 ^2- 1966 50 100 150 200 LENGTH fern) 500 Figure 62. Length composition of samples of bluefin lima taken in fuikey in 1967 and 1968 ("+" on graph means less than 0.5%). 53 cm) individuals in various locali- ties in the northeastern corner of the Mediterranean in November and De- cember. They also noted catches of young tuna off Haifa, Israel, in wa- ters with salinities of 39.58 o/oo or more. Sara (1964, Figure 1) indicated that the northeastern comer of the Mediterranean (Haifa. Israel, to the Gulf of Mersin, and around eastern Cyprus) is a year-around habitat of 5-25 kg bluefin and an August-No- vember habitat of 0.2-3 kg bluefin. The southeastern corner of the Mediterranean (coasts of Egypt and southern Israel is evidently devoid of bluefm tuna. The stranding of a single bluefm near Alexandria (Heldt 1937) was regarded as an extraordinary event. d. Summary. Sara (1964, 1973) summarized the distribution of bluefin tuna in the Mediterranean Sea. From August through November, almost all of the Mediterranean is filled with large schools of very young tuna born in the year. After November, when they have attained a weight of about 2 kg, few are taken, probably because of bad weather rather than because of their absence. With the return of good weather in late winter, they are found in practically the same places, weigh- ing 4-5 kg. These young tuna remain in the areas where they were born, making small movements in response to changes in environmental condi- tions, and in search of food, until they reach sexual maturity. Sara's (1964) distribution chart indicates scattered concentrations of bluefin weighing 5-25 kg near most of the islands and coasts in the western and central Mediterranean, except the Af- rican coast between Morocco and Tu- nisia. In the eastern Mediterranean, concentrations are shown in the Aegean Sea. around the Bosphorus. and off the northeastern coasts. Medium-sized tuna (30-100 kg) are found off many of the coasts throughout the year. Sara's (1964) chart locates these fish along the Af- rican coast from Gibraltar to Bengasi (mainly from Tunisia to Cape Misurata). along the Spanish and French coasts and around the Balearic Islands, in the central, eastern and southern Tyrrhenian Sea, along the 66 east coast of Sicily and around the islands in the Sicilian Channel, in the Adriatic Sea and in the northwestern Aegean Sea. Sara (1973) said that these fish are especially visible in late summer and autumn, when they chase prey relentlessly on the surface near the coasts. Large tuna (over 150 kg) are abundant, except in certain local ar- eas, only from April to September (Sara 1964, 1973). The principal ar- eas of abundance indicated in Sara"s (1 964) chart are off western Sardinia, Sicilv, Tunisia and western Libva. Other occurrences are shuwn near Gibraltar, in the Gulf of Lion, in the Bosphorus, and off Bengasi. Sara (loc. cit.) believed that the major oc- currence of large bkiefin in the Medi- terranean is the spawning run from the Ibero-Moroccan Bay, whose vol- ume and movements are greatly in- fluenced by the inflowing Atlantic Current. Occurrences in other sea- sons consist of relatively small num- bers offish which stay in areas where food is plentiful, such as the Gulf of Lion (off the mouth nf the Rhone), the Tuscan Archipelago, the Aeolian Islands, the Strait of Messina and the Dardanelles. I'lie fisheries for large biuefin were historically the mo.st important in the Mediterranean, and laws pro- hibited the capture of young fish in Italian waters more than a century ago. Catches of young biuefin have increased since 1920, and this ten- dency has accelerated after World War II. Perhaps in consequence of this, the landings of large biuefin by the traditional trap method declined, especially in the 1970s. The intro- duction of longline and purse seine fisheries for large bluetln in this pe- riod, however, had restored the catches of large fish to their earlier level bv 1974. 67 V. SPAWNING A. INTRODUCTION Considering the amount of re- search which has been devoted to the Atlantic bluefin tuna, positive informa- tion on its spawning habits is surpris- ingly incomplete. ITie earliest and most extensive information on this subject is from the Mediterranean, but even there much remains to be learned. The situa- tion is far worse in the important east- em North Atlantic region, since, to our knowledge, no larvae or small (less than 12 cm) juveniles have been col- lected in any of its areas which have been assumed to be spawning grounds. Greater advances have been made in the western North Atlantic, but posi- tive information there is limited to rela- tively small areas. Aside from a few larvae collected near the Equator, wc found no information whatsoever on the spawning of the species in the South Atlantic. The paucity of knowledge con- cerning this vitally important phase of the life of the bluefin tuna may be attributed to the serious limitations in- herent in the methods used to obtain this information. B. DEFINITIONS Larva: Matsumoto et al. (1972) stated that the larval stage of nearly all tuna species ends when the larva has attained 1 to 13 mm standard length (SL) (Figure 63). We have arbitrarily selected 12 mm SL as the size at which the bluefin tuna passes from the larval to the juvenile stage. Juvenile: Postlarval fish may be regarded as juveniles until they reach maturity— a considerable size, in the case of the bluefin tuna. In this section on the spawning of this species, how- ever, we have arbitraril> limited the tenn "juvenile" to include individuals from 12 to 120 mm. Bluefin tuna only slightly larger than 1 20 mm are active predators and are taken in some fisher- ies. Index of Maturity: The numeri- cal index of maturity most frequently used in this section is the weight of the entire fish divided by the weight of the gonads. This index is therefore inversely proportional to the maturity of the fish. Other indices are defined where they first occur in the text. C. CRITERIA FOR DETERMINING SEASONS AND AREAS OF SPAWNING, AND THEIR LIMITATIONS 1. Presence of Ripe Adults The evaluation of the condition, or degree of maturity, of the gonads of adult fish is one of Iwo approaches used to determine the spawning sea- sons and areas of the bluefin tuna. The presence of fully ripe adults is assumed to indicate spawning. Stages of maturity have been evaluated by several methods, some subjective and some objective. The ap- pearance and physical characteristics of the ovaries and the ovarian eggs are among the subjective criteria used for this purpose. Visual diagnosis may be based on the color of the ovaries and the extent and color of their external veining. Their consistency (hardness or softness ) and their size and shape are Figure 63. Early life history stages of TImnmis ihynnus of the western central Atlantic (Richards 1089). 68 important physical factors. The size and consistency of the ovarian eggs, and the tenacity of their adhesion to one another and to the wails of the ovaries also provide indications of maturity. The more objective criteria include indices of maturity, usually expressed as a relationship between the weight of the ovaries and some measurement of the fish from which they were taken. Examination of histological sections of gonads permits more exact assessment of stages of maturity, but few such studies on bluefin tuna have been pub- lished. Estimates of spawning areas and seasons from gonad condition are sub- ject to many limitations. Unless the specimen is fully ripe, serious errors may be involved, as these fish can travel great distances in short periods (Mather 1962, 1969). Also, frilly ripe bluetln tuna are difficult to catch on hook and line as they are reportedly reluctant Sara to feed (Sella 1929a, Rivas 1954, Sara 1964, 1973). Their capture in traps is extremely rare. Sanzo ( 1 9 1 Oa) found a ripe female in a trap near Palermo, Sicily, in 1908, but had not encoun- tered another when he produced his final (1932) work on the eggs and lar- vae of the bluefin. Likewise Rodriguez- Roda ( 1 964a, 1 967a) had observed only one ripe female in his extensive studies at Barbate, the most productive tuna trap in this century, and other sites. He remarked that the oviduct of this indi- vidual might have been obstructed. The recently developed purse seine fishery in the central Mediterranean spawning grounds (Section IV) has made more mature bluefin available (R. Sara, per- sonal communication), but no studies based on these catches have appeared. Bluefin tuna spawn fractionally, by several separate emissions of eggs over a period of several days, rather than by a single emission (Sanzo 1910a, Frade and Mana9as 1933). Thus, even when ripe fish are taken, the localities of capture may not indicate the total area over which the fish have spawned. The fractional nature of their spawning also adds to the difficulties in determining when bluefin tuna are frilly ripe. As noted above, many of the criteria used to determine stages of maturity are subjective. In addition, even the seemingly objective indices of maturity may be subject to error. Maturing and post-spawning fish may have the same indices as the size of their gonads increases and then de- creases. Also, maturing fish are con- siderably heavier than spent ones of the same length. This fact introduces er- rors when indices for fish in these dif- ferent phases are compared. In addi- tion, the testes of males decrease in size much more than tlie ovaries of the fe- males after spawning has been com- pleted. Thus comparison of indices for fish of different sexes may be mislead- ing. Frade (1937) pointed out these sources of error and used corrections for the differences in gonad weight be- tween sexes and in total weight be- tween maturing and spent individuals of the same length. These refinements in the method, however, have appar- ently been ignored. We have observed another possible source of error in indi- ces of maturity for bluefin tuna in the western North Atlantic. When large bluefin tuna arrive in New England waters in July (Section IV), they are spent and their gonads are small. To- ward the end of their feeding season, in late summer and early fall, their go- nads become almost completely en- cased in adipose tissue. If this tissue is not removed before weighing the go- nads, the maturity index may be almost as high as for a ripe fish, even though the gonad proper is small and dormant. Baglin (1976) noted that the average weight of the adipose tissue attached to the ovaries of six giant bluefin taken off New England in August 1975, was 1,152 g, while the average weight of the gonads themselves was 1,1 14 g. Tlie well documented presence of fully ripe fish may therefore be a good indication of a spawning area, but con- clusions based on nearly ripe fish may be very misleading. 2. Presence of Pelagic Eggs and Larvae The times and locations of collec- tions of pelagic eggs and larvae form another basis for the detemiination. or estimation, of the seasons and areas of spawning. When eggs or extreme!) small larvae can be identified, it may be assumed that spawning occurred not very long before ihe time of collection and not ver>' far from the collecting locality. Estimating the time and place of spawning when larger larvae or ju- veniles are collected, however, requires knowledge of their growtli rate, and also of the currents in the collecting area. Although this approach is very logical, its execution has not proved to be easy. The collection and identifica- tion of eggs and larvae is difilcult. Blue- fin eggs cannot be positively identified unless they are hatched and the larvae are reared to an identifiable size (PiccinettiandPiccinettiManfrin 1970), which is a most difilcult process. The identification of larvae has been con- troversial (Sella 1924,Dieuzeide 1951, Duclerc et al. 1973, Richards 1976). Sella and Richards questioned Ehrenbaum's (1924) tentative identifi- cations of "Orcyniis ihynnus L.'\ and Richards reidentified several of Ehrenbaum's larvae of tunas and tuna- like species. Duclerc et al. ( 1 973) found it impossible to distinguish eggs and larvae oVThunnus ihynmus L.", from those oV'Aitxis lhuzard\ even after the eggs had been hatched and tlie larvae reared to lengths of 5.0 mm. Duclerc et al. (1973) and Richards (1976) ques- tioned Sanzo's (1932) descriptions of the egg and larval stages ofOrcyims t/iynnus Ltkn.", and felt that they should be verified. Some encouraging progress has been made recently, however. Duclerc et al. ( 1 973) and Scaccini et al. (1975) have reared larval bluefin tuna for periods of up to eight days from eggs collected at sea. Richards and Potth off (1974) have produced a most thorough description of larval bluefin tuna more than 3.0 mm (SL) long. The accuracy of estimates of spawning seasons and areas based on the collection data for young stages, their length and growth rates, and cur- rent systems in the collecting area, de- clines rapidly with the passage of time from the hatching of the specimen. Not only does the error in calculating pas- sive transport increase, but juvenile bluefin tuna become active swimmers at a surprisingl\ early age. Tlie\ attain their full complement of caudal rays at about ! 6 mm SL (PotthofiF 1 975). Sella (1929a) stated that the bluefin tuna could already be considered an active fish at the age of 1 5 days. Most of the available data on the growth of the early stages of the blue- fin (Section 111) have been presented in terms of weight. Growth data in terms 69 '■& Major Known Spawning Aieas '* Minor, Onavaluatad or Hypothetical Spawning Jlreas Figure 64. Bluefin spawning areas in the Mediterranean and Blaclc Seas and tlie eastern Atlantic. of length would be much more useful for back-calculating the localities where larvae or juveniles had probably been spawned. Collection data for eggs and small larvae which have been reared to iden- tifiable sizes furnish good indications of spawning dates and areas. Estimates based on collection data for larger lar- vae or juvenile stages, however, must be regarded only as approximations. Unfortunately, most of the avail- able data on the growth of early stages of bluefin tuna(d'Amico 1816, Bourge 1908, Heldt 1930, Piccinetti and Piccinetti Manfrin 1 970) were presented in terms of weight (a characteristic which is seldom reported in this size range), rather than length. Length-age data for early stages of bluefin tuna are needed to reduce the error in estimat- ing the probable date and location of hatching from the collection data of early stages. D. SPAWNING AREAS AND SEASONS 1. Mediterranean and Black Seas a. General Information The spawning habits of T. ihynnus thynnus have been more extensively studied and are better known in the Mediterranean than elsewhere. Bluefin tuna spawn over exten- sive areas of the Mediterranean and Black Seas (Figure 64). The larger in- dividuals, to which most of the research has been devoted, spawn mainly in the last half of June and the first half of July. There is considerable evidence that the smaller bluefin which have reached maturity (the older small blue- fin and the medium-sized fish) spawn later, throughout July and into August, and occasionally even into September. The principal known spawning ar- eas were in the southern central Mediter- ranean, around Sicily, off western Sardinia, and off northeastern Tunisia. Additional reproduction has occuned around the Balearic Islands, off Tripolitania (western Libya), off Alge- ria, and in the Black Sea. It is believed that additional research would show that the species spawns over a much broader area. The relationship between eastern Atlantic and Mediterranean bluefin tuna has been debated for centuries. An an- cient theory, proposed by Aristotle (circa 325 B.C.), has been believed for centuries by the Mediterranean fisher- men and repealed by many authors. His hypothesis was that the bluefin tuna lived in the Atlantic for most of the year, but entered the Mediterranean in the spring, travelled along its northern shores to the Black Sea to spawn, then returned to the Atlantic, following the southern coasts of the Mediterranean, in the summer. Scientific findings in the sixteenth and seventeenth centu- ries, however, raised questions about this theory and from 1 875 to 1 925 most authorities believed that spawning blue- fin did not pass through the Strait of Gibralter in numbers. They felt that the Mediterranean bluefin tuna comprised an autochthonous population, which made only limited local movements to and from the spawning areas. Since 1925 evidence of migrations from the Atlantic into the Mediterranean has re- opened the question. A recent hypothesis (Sara 1964, 1973) satisfies many of the arguments used by both sides in this debate. Sara maintains that most of the large bluefin which spawned in the Mediterranean were migrants from the Atlantic and spent the rest of the year in that ocean, but that most of the medium sized and small fish which spawned in the Medi- terranean had been bom in that sea and had remained there until they attained about 1 50 kg. These problems are dis- cussed in detail in Sections IV and VI. b. Specific Occurrences The first positive indication that bluefin tuna spawned in the Mediterra- nean was presented by Cetti (1777) from observations made along the coasts of Sardinia. He found eggs, which he believed to be those of blue- fin tuna, on the ropes of the traps, and also in the water around the traps. He also stated that the ovarian eggs offish taken in the traps in May were very well developed, but that those of fish taken in June had degenerated. He con- cluded tentatively that more large blue- fin tuna spawned in the Mediterranean than in the Black Sea. Even though his data are questionable, in view of present knowledge, his conclusions were cor- rect. His work eventually led to the first doubts about the Aristotelian theory. This was followed by d'Amico's (1816) revelation that bluefin spawned off Sicily. He noted that juveniles of the species weighed about 42 g in June, 122 g in August and 840 g in October. These data and those of Sella (1929a) enabled Heldt (1930) to present the first growth curve for the early stages 70 of this species. Sanzo (1909, 1910b) described eggs taken from ripe bluefin caught off Sicily. Later Sanzo (1929, 1932) described pelagic eggs which he collected in the Strait of Messina in May and June, and larvae which were hatched from these eggs and reared for six days. Sanzo concluded that these were the eggs and larvae of bluefin tuna, because the eggs matched those taken from a ripe female, and the six- day-old larvae hatched from them matched the smallest of the identifiable bluefin tuna larvae which he had col- lected off Messina. More recently, how- ever, several authors who have exam- ined Mediterranean material have con- cluded that bluefin tuna eggs could not be identified positively unless they were hatched and the larvae were reared to an identifiable size (Piccinetti and Piccinetti Manfrin 1970, Duclerc et al. 1 973, Scaccini et al. 1 975). Collections of larger larvae and juveniles were also reported (Sella 1924, 1929a; Sparta 1933, Padoa 1956, Scaccini 1965, Piccinetti and Piccinetti Manfrin 1970). Although Sella (1924) explained the differences between juvenile T. rliynnus Ihynnus, T. alalimga and Auxis bisus, these early collections were not thor- oughly described or illustrated. Recently Scaccini (1961) and Scaccini et al. (1975) published a photograph of some of Sella's material and reproduced drawings of larvae 4-12 mm long pre- pared under his guidance as well as their own photographs of eggs, larvae and juveniles. Watson identified and described, with illustrations, larval and juvenile bluefin from 11.4 to 37.8 mm long collected off Messina, Sicily, during the summer months of 1958, 1959 and 1 960 (Watson and Mather 1 96 1 ). These identifications, based on external char- acters (Sella 1924) and osteological characteristics examined by radiogra- phy, were subsequently verified by T.C. Potthoff (personal communication) of the Southeast Fisheries Center of the National Marine Fisheries Service, who examined additional vertebral charac- teristics after clearing and staining the specimens. Many additional identifications of juvenile Mediterranean bluefin tuna were obtained after the Woods Hole Oceanographic Institution collection was donated to the Southeast Fisheries Center of the National Marine Fisher- ies Service at Miami, Florida. TC. Potthoff (personal communication) identified 60 of the specimens which had been collected in Sicilian waters during the months of July, August and September as Thunmistlwrmusthyiimis. The lengths and the dates and localities of capture of these specimens, which include the four largest ones reported by Watson and Mather (1961), are shown in Table 16. The most numerous records of lar- val and juvenile bluefin tuna are from Sicilian waters, but these forms have been encountered in several other parts of the Mediterranean. Roule (1917) cited Bourge's (1908) records which showed the ap- parent growth of juvenile bluefin col- lected off Tunisia. These young fish weighed about 150 g in August, at- tained 900 g in September, and ex- ceeded 1 kg in October. Richards (1976) reidentified two of Ehrenbaum's (1924) larval speci- mens, originally identified as ''Eulhynniis alliteratus Raf (?)", taken near the Balearic Islands in late August 1910, as Thunnus thynnus ihynnus. Roule (1917) had tentatively listed these islands as a spawning area for this spe- cies. In 1972 Duclerc etal. (1973) col- lected eight bluefin tuna larvae near the Balearic Islands between June 20 and July 7. They also collected pelagic eggs which fitted Sanzo's description of those of bluefin tuna, but, even after hatching some of these and raising the larvae for periods of up to four or five days, they were unable to distinguish eggs of T ihynnus ihynnus from those of ".4!tY/.v ihazard\ Dieuzeide (1951) listed the cap- tures of three bluefin tuna larvae off the coast of Algeria June 7-8, 1940. He provided drawings of two of these, which were 5 and 6 mm long. Table 16. Collection data and lengths of bluefin tuna from Sicilian waters. M.E. Watson originally identified most of the specimens. T.C. PotthofT subsequently confirmed her determinations and identified the remaining individuals. Collection Area Collection Date Standard Length (mm) I. Straits of Messina July, 1949 July 15 -Aug 31, 1958 July- August, 1959 June 30 -July 1, I960 July 18-20, 1960 July 24 -25, 1960 July 26 -27. 1960 August 20 -21, 1960 August - September, 1 960 August 10-20, 1963 Unknown 2. Palermo August 1963 17.4, 19.4,24.9,34.3.36.8,37.1 33.7 30.5,31.0,31.0,32.0,34.2 16.9 17.0, 17.7, 19.5,22.5,24.8,25.1,26.3.26.9 1 7.7, 1 8.9, 20.9, 2 1 .0, 25.9, 28.0. 28.9 32.4,33.4,34.5,38.5 45.3, 50.4 50.5 17.6 62.0, 64.6, 66.4, 7 1 .3. 74.0, 76. 1 . 87.5. 99.6 12.6, 15.5, 16.7 16.9, 16.9, 17.1, 17.3, 18.7. 19.5. 20.1, 20.3.20.5,27.3,30.4 107.8, 117.5 The Messina specimens were purchased by the Woods Hole Oceanographic Institution from G. Arena of Messina. I he Palermo specimens were donated to the Institution by P. Tarantino of Dorclicslcr. Mas.sachusctts. .All ofthc specimens ucrc subsequently donated to the Southea.st 1 Ishcries Science Center. National Marine Fisheries Service, Miami, Florida, and arc now in its Lollection. Recently, Piccinetti (1973), Piccinelti and Piccinetti Man frin ( 1 973 ), and Piccinetti et ai. (1976b) reported on occurrences of larval and juvenile Thunnus thynnus thynnus in the Adriatic Sea, the first positive indication that the species reproduced there. Richards ( 1 976) thought that a lar- val specimen collected off Cape Matapan, Greece, on August 26, 1911, and originally identified by Ehrenbaum (1924) as "'Orcynusgermo Lacep. (?)" was probably Thunnus thynnus thynnus. Richards agreed with Sella ( 1 929a) that the specimens identified by Ehren- baum (1924) as '"Orcynus thynnus L. (?)" were not of that species. Several scientists have reported on eggs and larvae of bluefm tuna col- lected in the Black Sea in mid and late summer. Vodyanitskii and Kazanova (1954) reported that eggs and larvae were encountered there repeatedly in the second half of the summer, but mainly in the beginning of August. Vodyanitskii (1936) and Oven (1959) described eggs found in the latter part of July and in early August, off Sevastopol and Karadag, respectively. Oven hatched some eggs and reared the larvae for up to eight days. Many authors have presented in- dices of maturity for bluefin tuna taken in the Mediterranean traps; we cite those of Sara (1964, 1973) because they are among the most recent. He found that the first maturing fish taken in the Si- cilian "arrival" traps in early May had maturity indices of 70-60, whereas those taken toward the end of this fishery in mid Jime had indices of about 25. An index of about 40 was the most fi-e- quent within the entire period. The fe- males were generally in a less advanced state of maturity than the males in this period. During the "return" run of pre- sumably spent fish, the first captures, in eariy July, were of fully mature fish with indices of about 50. By late July, this had increased to 9 1 , and in early August, to 198. This decrease in matu- rity was much more regular than the increase during the "arrival" run. Sella (1929a), as well as Scordia (1932) and Sara (1973), noted that small maturing or ripe fish were sometimes taken in the Sicilian return traps, along with the spent large ones, in July and even in August. Sella and Sara also pointed out that the "arrival" traps in Tripolitania took relatively small maturing fish for a considerable period after the conclu- sion of the "arrival" fisheries for larger maturing individuals in the Italian and Tunisian traps. Heldt (1938) summarized the re- sults of Frade's histological studies of gonads of bluefin tuna taken in various Tunisian traps. Heldt had collected this material and furnished it to Frade. Frade concluded that, if this small sample was representative, the various groups of tuna which frequent the Mediterra- nean do not mature simultaneously. The fish taken off Tunisia were, in general, less mature than those taken in the east- em Atlantic at the same time. Gonad studies related to the spawn- ing of bluefin tuna in tlie Black Sea depended mainly on material collected near Istanbul, Turkey. Hovasse ( 1 927), af^er analyzing statistics of landings at that city in 1915 and 1921-1923, found two significant maxima in the catches, one in the spring (March-April) and one in the summer (July). There was also a minor maximum, whose signifi- cance Hovasse did not understand, in the winter (December). He believed that the spring peak corresponded to a northward migration into the Black Sea and the summer one, to a southward passage back from the Black Sea into the Sea of Marmara. He found that the ovaries of the females taken in the spring were full, whereas those taken in July contained no eggs. Akyiiz and Artuz (1957) stated that bluefin started to run through the Bosphorus into the Black Sea in April, and that this run peaked in July and ended in September. The re- turn migration occurred from late Oc- tober to November. Their studies of gonad condition indicated that spawn- ing occurred in late July, August, and possibly September, ly igiingor ( 1 957), however, listed three periods of pas- sage: November-January, March-April, and July-autumn. Thus there are con- siderable differences in the interpreta- tions of the catch records in regard to migrations and spawning. The above research shows clearly that spawning in the Mediterranean by large bluefin tuna starts about mid-June and ends about mid-July. The repro- duction of the smaller (the medium- sized and some of the larger "small" individuals) fish has not been investi- gated as thoroughl) . The a\ ailable evi- dence indicates that tiiey begin to spawn later than liie large ones, and that their spawning extends through July well into August, and sometimes even into September. This information is based mainly on observations made in the south-central Mediterranean. The spawning periods in other parts of the Mediterranean and in the Black Sea, are not as clearly defined. Reproduc- tion off the Balearic Islands evidently extends from mid-June at least to late August. Collections of eggs and larvae indicate that spawning in the Black Sea is at its maximum in the second half of July and eariy August, but conclusions based on catch records and gonad con- dition are inconsistent. The most completely documented bluefin tuna spawning areas in the Meditertanean are off northeastern and western Sicily, and off the Balearic Is- lands. Another apparently well docu- mented area is in the Black Sea. There are also strong indications of spawning off western Sardinia, northern Tunisia, and Tripolitania. More fi^gmentary data suggest reproduction in the Adriatic Sea, off Algeria and Greece, and in the Aegean Sea. Scaccini et al. ( 1 975) con- cluded that additional intensive research will show that the spawning areas of T. thynnm thynnus cover much more of the Mediterranean than was then be- lieved. 2. Eastern and Central North Atlantic and the South Atlantic a. General Information Many experts have assumed that the Ibero-Moroccan Bay (west of Gibraltar) contains the major, or the only, spawning grounds for T thynnus thynnus in the eastern Atlantic. Exten- sive research in this area, dating from the late nineteenth century, has pro- duced much information on the condi- tion of the gonads of the bluefin tuna taken in the once flourishing trap fish- eries in this bay. As noted in Section IV, this fishery was divided into nvo periods, the "arrival" (late April, May and June) and the "return" (July and August). During the "arrival" period, the fish are apparently travelling along both coasts of the bay in the general direction of the Strait of Gibraltar. In the "return" fishery, which occurs only along the Iberian coast, they seem to be travelling westward. The "arrival" traps which have been used within recent years were distributed from Cabo de Santa Maria in Portugal to Tarifa, Spain, and along the Moroccan coast from Kenitra to Cape Spartel, Morocco (Fig- ure 44). It should be noted that the two latter localities are at the very entrance of the Strait of Gibraltar. Some of the Portuguese and Spanish traps were re- arranged at the end of June to fish in the "return" period. The gonad studies showed that the "arrival" fish were maturing and that their indices of maturity became lower (indicating more advanced maturity) as the season progressed. On the other hand, the "return" fish were generally spent, with their indices of maturity increasing (indicating shrinking of the gonads) toward the end of the season. Since the more mature ovaries were an important by-product of the fishery (F. de Buen 1 928), these facts were known to the fishermen for centuries before the scientists arrived on the scene. In fact, they may have played a role in the development of Aristotle's migratory theory. This simple combination of cir- cumstances was interpreted in opposite ways by the proponents of the "Atlan- tic-Mediterranean migration" or "mi- gratory" theory and the advocates of "separate Atlantic and Mediterranean populations" or "sedentary" theory. The former maintained that these facts indi- cated that many of the "arrival" tuna entered the Mediterranean to spawn, and that the "retum" run included many individuals which had spawned in the Mediterranean and were returning to feeding areas in the Atlantic. The latter believed that few, if any, of the "ar- rival" fish passed through the Strait of Gibraltar and that most of them spawned in the Ibero-Moroccan Bay. Despite the many decades of in- tensive research on bluefin tuna in the area, however, no identifiable eggs, lar- vae or juveniles (as defined here) of bluefin tuna have been collected in tlie Ibero-Moroccan Bay. In addition to the gonad studies, a great deal of other biological informa- tion pertinent to the possible spawning of the species in this area has been published. Extensive data on the prop- erties of the environment are also avail- able. Evidence of reproduction of the species in other parts of the eastern Atlantic, and in the central and the South Atlantic (Figure 65), is ver> meager. Some larvae have been collected in the eastern equatorial Atlantic. A few oc- currences of ripe bluefin in the Bay of Biscay have been reported. Spawning in the vicinity of the Azores has been suggested. A spawning area between Lanzarote, one of the Canary Islands, Conception Bank, and the Moroccan coast has been hypothesized, but not investigated. b. Specific Occurrences From numerous studies of the con- dition of the gonads of bluefin tuna caught in the Ibero-Moroccan Bay, we have selected for discussion those of Frade and Manai;as (1933) and Frade ( 1 937a) for Portugal, F. de Buen ( 1 927) and Rodriguez-Roda ( 1 964a) for Spain, and Lozano Cabo (1957, 1958) and Fumestin and Dardignac (1962) for Morocco. Frade and Mana^as (1933) pre- sented one of the few histological stud- ies on the stages of maturity of the gonads of bluefin tuna which are avail- able. Their observations confirmed the fractional nature of the spawning of the species. This accounts for the lack of completely ripe fish in the traps, be- cause all the eggs destined to be emit- ted in a given season do not mature simultaneously. A large number of the individuals entering the traps in the "arrival" period may have already emit- ted part of their spawning products. In addition the small bluefin I m long (about 3 years of age) behave differ- ently according to their sex. The males are in active or terminated spermato- genesis, whereas the females are in very retarded ovulation. Later Frade (1937a) described a simple method for determining stages of maturity consistent with the abso- lute ones established by his histologi- cal studies (Frade and Mana^as 1933). To offset the effect of seasonal varia- tions in the length-weight ratio, his in- dex of maturity was the quotient of the weight of the head, rather than the total weight ofthe fish, divided by the weight of the ovaries. He also used a factor K (the ratio ofthe weight of ovaries to the weight of testes at the same stage, as determined by histological studies, usu- ally about 1.5, to make the indices for males comparable to those for females at the same stage of maturity. Even though these corrections did not pre- vent the occurrence of similar indices, in the intermediate stages, for pre- spawning and post-spawning fish, he felt that this procedure clearly showed the progressive emptying of the go- nads, or the fractional emission ofthe spawning products. His stages are shown in Table 17. Frade (1937a) presented the indi- ces obtained for 171 bluefin tuna ex- amined at Vila Real de Santo Antonio, Portugal, from May 31 to August 18, 1933 (Table 18). These same values are shown graphically in Figure 66. Frade leached the following conclu- sions: 1) Ripe and intermediate stages (A- D) of both sexes are represented in June, but in decreasing frequen- cies. 2) In July intermediate stages (C and E) predom inate among the females, and in the males, late intermediate and spent stages (E and F) appear. 3) In August intermediate stages (D and E) are predominant for the fe- males while ripe and intermediate stages (B and C) are still pre- dominant among the males. Some fish of both sexes have attained the fully spent stage (G) and females in the fiilly ripe stage (A) have disappeared from the catches. F. de Buen ( 1 927) noted that blue- fin tuna which had arrived to spawn were caught in the southern Spanish Atlantic traps in April, May and June, and that they also entered these traps beginning in July, after spawning. Af- ter examining a large number of speci- mens from these traps in 1923, he con- cluded from the data that in 1923 the bluefin spawned in the Ibero-Moroc- can Bay in the months of June and July, and that the males had attained sexual maturity before the females (Table 1 9). In one of his major works on the biology ofthe bluefin tuna taken in the traps off the southern coast of Spain, Rodriguez-Roda (1964a) provided much infonnation on the variations in gonad condition through the spawning season. He set up the following scale 73 ■^^■^-yy Figure 65. Bluefin spawning areas in the Eastern Atlantic and the Mediterranean and Black Seas. for stages of maturity, based on the color of the ovaries: I. Immature: Gonads of rosy color. II. Prematuration: Male gonads of a violaceous color, and female of a rose color. III. Maturation: Male gonads of a rose or a rosaceous white color, and female yellow rose or yellow- ish. IV. Prespawning: Male gonads of a milky violaceous color. V. Spawning. VI. Postspawning: The male and fe- male gonads are of a violet color. Very swollen ovaries with most of the eggs transparent were considered to be in stage V, Only one such female was observed during the "return" pe- riod and it was concluded that its ovi- duct was probably obstructed. During the "arrival" run in May and June the bluefin tuna appeared with turgid gonads in a state of prematuration or maturation and in the "return", in July and August, the gonads were al- ready flaccid and with obvious signs of having emitted the eggs. Between these two phases, at the end of June and early in July, when spawning should actu- ally occur, the bluefin disappeared from the coast and its captures were very limited. To define the spawning period, the monthly percentages of the sexual stages for 779 males and females ex- amined at the Barbate trap from 1956 to 1959 are shown in Table 20. The majority of the fish were in a prespawning state (stage III) in May and June and in the post-spaw ning stage (stage VI) in July and August. The small percentage of tuna ol'both sexes which were in stages I and II were not large individuals. Those in stage I in June were ^5 and 80 cm long, res- pectively. Those of both sexes in stage II were 1 15 to 140 cm long, excepting two females 170 and 175 cm long. In July and August a few bluefin were in the prespawning state. One male 135 cm long was in stage II in July. Males and females from 125 to 220 cm long in July, and 125 to 135 cm long in August, were in stage III. The onl\ specimen in stage IV was a male 170 cm long taken in June. Rodriguez-Roda ( 1 964a) also used an index of maturity, the "gonosomatic relation", equal to 100 times the ratio of the weight of the gonads to the total weight of the fish, to evaluate relative fecundity. He also provided plots and formulae showing the relationships be- tween the weight of the gonads and the fork length of the fish for bluefin tuna during the "arrival" and "return" peri- ods. He tabulated the gonosomatic re- lation for males and females, by 5-cm length groups and 10-day periods, for 260 "arrival" and "return" fish exam- ined in 1956, and also for 198 exam- ined in 1958 (Tables 21a and 21b). This index generally increased with size of fish. Its mean value also in- creased during the "arrival" period in June, for both sexes, then descended abruptly in the months of July and Au- gust, in the "return" period. All this confirmed spawning in early July. In general, the index was higher for fe- males than for males. The condition of the gonads of bluefin tuna taken off the Atlantic coast of Morocco was examined by Lozano Cabo (1957, 1958) and Furnestin and Dardignac ( 1962). Lozano Cabo( 1958) tabulated, by sex and 10-day period, the minimum, mean, and maximum indices of maturit}' of samples of blue- fin tuna taken in 1955 by the trap at Los Cenizosos (near Larache) in 1^)55 (Table 22) The mean index for females de- creased from 80.8 in the last ten days of May to values between 46.2 and 40.0 in the three ten day periods in June. In the second ten days of May a few fe- males with very low maturity (indices averaging 105. 3) were examined. When these fish were being taken, the catch was very small. The fisher) did not 74 Table 17. Indices of maturity and condition of gonads (Frade 1937a). Indices of Maturity Condition of Stage Female Male Gonads A Up to 7 Up to 10.5 Full B 7-12 10.5- 18.0 i» C 12-17 18.0-25.5 II D 17-22 25.5-33.0 Intermediate E 22-27 33.0-40.5 " F 27-32 40.5-48.0 Empty G 32-37 48.0-55.5 II Table 18. Stages of maturity for 1 7 1 bluefin tuna examined at Vila Real de Santo Antonio. Portugal. May 31-August 18, 1933 (Frade 1937a). FEMALES A B C D E F G 29 Tuna 13 11 4 1 — — — May 3 1 -June 5 % 45 38 14 -> — — — 21 Tuna 3 5 7 4 2 — — % 14 24 33 19 10 July 7-18 54 Tuna — 3 10 15 14 6 6 % — 6 19 28 26 11 11 August 5-18 MALES A B C D E F G 16 Tuna 9 3 2 2 — — — May 31 -June 5 % 56 19 13 13 — — — 17 Tuna 3 8 4 - 1 1 - July 7-18 % 18 47 24 6 6 — 34 Tuna 1 8 9 4 6 3 3 Aueust 5-18 % 3 24 26 12 18 9 9 60 50 >40 ^30 10 FEMALE MALE ■ Jl '\ JI .N^ s. \ ■\ A / v \ ,> v> <•■■■'. / / ••^•. JU -'' \" '•■• JULY* ■ •-Vv ..«.. AUG« 1 \ auq4 •'1 -• A B C D E F G F G Figure 66. Plot showing the changes in sexual maturity of Atlantic bluefin tuna by month. The data are the same as in Table 18. become productive until the index reached 63.6 (LozanoCabo 1957). The maturity of the females decreased slightly (index rising fi-om 40.0 to 43.9) from the second to the third ten days of June. This was caused by the capture of some spent, post-spawning, or "return" fish in this "arrival" trap. The only in- dices for males, taken in the last ten days of May, ranged from 42.7 to 1 3 1 .2. with a mean of 76.8. These figures do not differ greatly from those for fe- males taken in the same period. Fumestin and Dardignac (19^2) tabulated the mean "gonado-somatic indices" (evidently the "gonosomatic relation" ofRodriguez-Roda 1964a) for males and females taken in the trap at Moulay-bou-Selham (near Kenitra) in the first and second halves of May and the first half of June (Table 23). The indices for males increased rather regularlv from 1.10 in the first half of May to 2.04 in the first half of June. Those of the females followed a similar trend, from 1.52 to 1.99 over the same period. TTiese values showed a more consistent increase in maturity through the season than those of Lozano Cabo ( 1 958). Fumestin and Dardignac ( 1 962) also plotted the gonado-somatic indices for each sex against the length of fish. Those for females increased with length, but those for males only increased up to a length of 210 cm and decreased slightly in longer individu- als. The authors attributed this to the fact that the largest fish were caught at the beginning of the season, when their mauirit)' would normally have been less advanced. The maturity of the fish taken in the first half of June at Moulay-bou- Selham is generally comparable to that found by Rodriguez-Roda ( 1 964a) for fish taken at Barbate, Spain, in the first twenty days of June. Lozano Cabo ( 1 958), however, reported that the blue- fin tuna which he examined at Los Cenizosos were less mature than those which he had examined at Barbate. I'liere is sn^ong circumstantial evi- dence that bluefin tuna spawn in the Ibero-Moroccan Bay, but conclusive proof of this is lacking. Great numbers of maturing individuals occur there in May and June, and recently spent fish are abundant in July and August. The stages of maturity of these fish have been verified by many investigations. On the other hand, much research and 75 Table 19. Bluefin maturity by month from the trap fishery along the southern Atlantic coast of Spain (F. de Buen 1^)27). Males Females Month % Not % Fully % Mature Mature Spent %Not % Fully % Mature Mature Spent May 100 0% 100 June 32 60 8 67 30 3 July 1 33 66 26 12 62 August (beg.) 100 100 two intensive surveys have failed to produce a single identifiable egg or early stage (less than 1 cm long) of the species from the area. O. de Buen ( 1 924) and F. de Buen (1924, 1925. 1927) failed to find any eggs or early stages of bluefin tuna during their intensive 1923 survey of the Ibero-Moroccan Bay, west of Gibraltar, and the Alboran Sea, east of it They explained this apparent contra- diction of their theory that the Ibero- Moroccan Bay was a major spawning ground for bluefin tuna, and that the maturing fish did not enter the Medi- terranean to spawn, as follows. The bluefin spawned in this Bay, but the eggs and larvae were carried from it by the surface (Atlantic) current into the Mediterranean before they attained full mobility. There they accumulated off the Moroccan coast between Punta Almina and Cape Tres Forcas, where the rich plankton provided food for the early stages. Roule's (1923) contrary view, that bluefin larvae were abyssal and were passively transported from the Mediterranean into the Atlantic by the deep outflowing (Mediterranean) current will be discussed in Section VE2. More recently, Spanish scientists in the research vessel "Comide de Saavedra" conducted an intensive planktonic and hydrographic survey of the waters east and west of Gibraltar in 1972. The period of the survey, 19 June- 16 July, was chosen because this was regarded as the period of maxi- mum spawning of bluefin tuna in the area (Rodriguez-Roda 1975). During this cruise 66 stations were occupied in the Atlantic, four in the Strait of Gibraltar, and 82 in the Mediterranean (Rodriguez-Roda 1975). No identifi- able larval or juvenile bluefin tuna were collected. Until eggs or early stages collected in the Ibero-Moroccan Bay are posi- tively identified as those of bluefin tuna, this area must be regarded as only a hypothetical spawning area forthis spe- cies. F. de Buen (1937) proposed a sec- ond spawning area for the eastern At- lantic bluefin in the southeastern cor- ner of the Bay of Biscay. This hypoth- esis was based on reports of bluefin with ripe ovaries being taken there in June. Le Gail (1952) reported similar findings there during the 1 950 and 1 95 1 seasons. Creac'h (1952) described the sizes of bluefin tuna landed at Saint- Jean-de-Luz through the 1 95 1 season (April I2-October 26) and the condi- tion of their gonads. The ovaries of very few specimens among the bluefin landed in the period June 16-30 were ripe. The fish landed in this period were in the 15-25 kg weight range. Creac'h (1952) noted that a similar situation had been observed for the first time, in about the same period, during the 1 950 season. No observations of mature fish during the remainder of the season were reported. Cort et al. (1976) and Cort (1977) provided the first detailed information on the gonads of maturing bluefin taken in the Bay of Biscay. These authors examined four bluefin 1 25- 1 49 cm long taken in the Bay of Biscay June 25-26, 1976, and seven 111 to 161 cm long taken there on July 15. 1976. On the basis of color (Rodriguez-Roda 1 964a), they estimated that the two males in the first group were fully mature, whereas one female was in pre-spawning con- dition and the other specimen was spent. They classified three females in the second group as pre-spawning and two as spent, and considered one male to be fully mature and the other spent. Cort et al. (1976), using the tech- nique ofRodriguez-Roda( 1967), mea- sured the diameters of eggs from two females. 148 and 152 cm long, of the second group. The diameters for the 148 cm specimen ranged from 0.10 to 0.66 mm, with many from 0.22 to 0.58 mm and the highest mode at 0.50 mm Those from the other ranged from 0. 14 to 0.58 mm. with most between 0.30 and 0.53 mm and a wide mode be- tween 0.40 and 0.46 mm They classi- fied these fish as in stage IV (pre-spawn- ing) or between stage 111 (maturation) and stage IV, since the diameters were greater than those of the ova and oo- cytes of the specimens placed by Rodriguez-Roda (1967) in stage 111. Cort (1977) noted that these observa- Table 20. Monthly percentage of the sexual stages (years 1956-1959) of tuna at Barbate (Rodriguez-Roda 1964a). Month Sexual Stages of Males. Number = 284 II III IV VI Sexual Stages of Females. Number = 495 I II III IV V VI May — — 100.00 — June 0.62 3.09 95.06 — July — 1.85 5.55 1.85 August — — 3.17 — 1.23 90.74 96.82 n.4i — 100.00 3.26 96.33 — 9.46 1.06 90.54 98.94 76 Table 21a. Variation of gonosomatic relation* with length offish and period of capture for male and female bluefin tuna taken in the Barbate trap in 1 956 (Rodrlguez-Roda 1 964a). Fork Length: 95-99.5 100-104.5 105-109.5 110-114.5 115-119,5 120-124.5 125-129.5 130-134.5 135-139.5 June MO June 11-20 June 21-30 July 10-20 July 20-30 June 1-10 June 11-20 June 2 1-30 July 10-20 July 20-30 0.11 Males (Arrivals) Males (Returns) Females (Arrivals) 1.14 Females (Returns) 0.55 1.37 0.33 1 44 093 1.55 1.19 . 1.16 051 Fork Length: 140-144.5 145-149.5 150-154.5 155-159.5 160-164.5 16.5-169.5 170-174.5 175-179.5 180-184.5 Males (Arrivals) June l-IO 1.30 - - 1.35 1 89 - 1,21 June 11-20 1 28 - - 1.66 080 1.19 June 21-30 096 - • Males (Returns) 2.00 0.97 July 10-20 - - - - - 69 July 20-30 - - - 1.00 Females (Arrivals) 54 084 June 1-10 1.36 080 1.39 1.46 1 18 096 June 1 1 -20 1.38 1.71 1.33 2.03 109 1,38 1,33 June 21-30 - - - 1 90 Females (Returns) 1.71 - July 10-20 - - - 0,98 1 07 0,84 July 20-30 0.73 080 - 0.74 69 078 0,75 1,18 2 28 0,71 2,50 0,51 79 82 Fork Length: 185-189.5 190-194.5 195-199.5 200-204.5 205-209.5 210-214.5 215-219.5 Mean Males (Arrivals) June MO - 1.37 1.62 1 13 1 45 June 11-20 - 1.43 2.48 1,71 2,28 June 21 -.30 1.22 2.57 1 96 Males (Returns) July 10-20 I 17 0.69 July 20-30 0,63 0.44 " 0.57 0.66 Females (Arrivals) June 1-10 1 80 1.53 1.25 2,06 1 14 June 11-20 - 1.85 1.01 2 U 1 (•') June 21-30 1,54 2.02 2.07 Females (Returns) July 10-20 1.03 1.06 0.94 0<)^ 79 July 20-30 0.76 0.82 075 79 85 I 91 I 71 1,24 082 20 1,31 1 22 22 1 48 1,31 10 1,63 . 5 089 - 16 0.63 1 83 37 1 53 1 55 36 1 51 - 24 1 71 - 26 0.92 64 0,77 'Maturity index obtained by multiplying the ratio of Ihe weight ol ihe gotiads ic the total weighl of the fish hy 100 (Rodriguez-Roda 1964a) 77 Table 21b. Variation of gonosomatic relation* with length offish and period of capture for male and female bluefin tuna taken in the Barbate trap in 1958 (Rodriguez-Roda 1964a). Fork Length: 120-124.S 125-129.5 I30-I34.S 135-139.5 140-144.5 145-149.5 150-154.5 155-159.5 160-164.5 June MO June 11-20 June 21-30 July 20-20 June 1-10 June 11-20 June 21-30 July 20-30 Aug l-IO Aug l-IO 0.95 2.67 1.54 0.77 Males (Arrivals) 190 148 154 2.58 1.92 Males (Returns) Females (Returns) 0.77 1.29 0.81 1.6.1 0.69 1.39 - - 0.87 - - Females (Arrivals) 0.60 052 1 04 2.47 1.25 0.72 2 1.' 2.02 1.10 178 170 1.26 1 40 1.62 1.04 0.80 Fork Length: 165-169.5 170-174.5 175-179.5 180-184.5 185-189.5 190-194.5 195-199.5 200-204.5 205-209.5 June l-IO June 11-20 June 21-30 July 20-30 Aug 1-10 June l-IO June 11-20 June 21-30 July 20-30 Aug l-IO 1.10 0.66 2.14 1.37 067 Males (Arrivals) 1.32 1.30 2.41 1 69 1.38 1 22 2.02 - - 1 ,73 2.06 1.87 Males (Returns) - - - 0.80 063 - - 1.00 0.68 055 Females (Arrivals) 081 44 1 95 - 1.50 2.1.1 1 46 1 .33 1 60 1 64 1 34 1.75 2 05 1 75 1.53 Females (Returns) 1.18 090 1 19 1.00 1 16 - 0.80 0.92 . 084 069 088 - I 86 0.60 1 00 3.14 Fork Length: 210-214.5 215-219.5 220-224.5 225-229.5 230-234.5 235-239.5 240-244.5 245-249.5 250-254.5 Mean Males (Arrivals) June l-IO 2.51 2.13 1.30 - 1 99 - June 11-20 1.07 2.25 1.73 - - - - June 21-30 • 2.64 2.76 " Males (Returns) - ' July 20-30 - - 0.64 0.90 0.72 47 0.61 Aug 1-10 0.42 - - - Females (Arrivals) 0.48 0.47 June l-IO 1.98 1.89 2.11 - - - - June 11-20 2.49 2.47 - 2.17 - - June 21-30 - - - - 1.94 Females (Returns) " " July 20-30 - 0.72 0.74 0.89 0.94 - - Aug 1-10 - - - - - - - 3.06 31 1 73 15 1 91 6 2,87 12 0.72 13 059 41 1.65 38 1 68 1 1 94 17 0.96 23 89 ♦Maturity index obtained by multiplying the ratio of the weight of the gonads tn llic lotal weight of the fish by 1 00 (Rodriguez-Roda 1964a). 78 Table 22. Indices* of maturity for 140" bluefin tuna taken in the Los Cenizosos trap (near Larache. Morocco) in May and June 1955 by sex and 10-day period (Lozano Cabo 1958). Indices for Females Min. Avg. Max. Indices for Males Month Min. Avg. Max. May 20-30 37.1 80.8 142.8 42.7 76.8 131.2 June 1-10 25.4 46.2 62.7 — — — June 11-20 30.0 40.0 52.0 — — — June 21-30 — 43.9 — — — — ■ The index of maturity is the ratio of the total weight of the fish to the weight of its gonads. " Seventy-six were examined individually in the first three periods. The index for the la.sl period was obtained by dividing the total weight of 64 fish by the total weight of their tions were too few to be conclusive, but hoped that future research might determine the possible existence of a bluefin spawning ground in the Bay of Biscay. Aloncle (1964) found that many 30-60 kg bluefin wintered between the Canary Islands and Morocco and hy- pothesized that one group of them spawned between Lanzarote Island (Ca- naries), Conception Bank, and the Mo- roccan coast. Distributional data sup- port this hypothesis, but studies of go- nad condition and collections of eggs and early stages are required to verify it. Ferreira (1932) reported that ripe bluefin were sometimes landed in the Azores in spring, but his statistics indi- cated that the species was not abundant there. The only reports of larval bluefin taken in the eastern Atlantic are from equatorial waters off Africa. Fourteen small larvae (mostly 2.5-4.5 mm long) were taken in the area bounded by lati- tudes 6°N and 8°S and longitudes 0° and 1 5°30' W. One of these was col- lected in February, eight in March and five in August (Richards 1 969, Richards and Simmons 1971, unpublished NMFS data reports for Geronimo cruises 3, 4 and 5). The surveys in which these collections were made were carried out in the periods February- April and August-October, missing most of the period in which the eastern Atlantic bluefin are believed to spawn. The oceanic distribution of bluefin tuna (Wise and Davis 1973) in the first quar- ter of the year shows concentrations of bluefin near where these larvae were found, but the distribution in the third quarter does not. It would be desirable to conduct a survey in this area in the second quarter, and also to examine the gonads of bluefin caught there in the months of February through August. Much remains to be learned about the spawning of bluefin in the eastern Atlantic, and about the possible role of the Mediterranean, both as a spawning ground and a nursery area for bluefin tuna from the eastern Atlantic. The de- gree of mixing of adults and new-bom Atlantic and Mediterranean tuna, if they are indeed separate, is another impor- tant matter about which there is no quantitative information. In view of the possible passive transport of spawning products from the Ibero-Moroccan Bay into the Mediterranean, the technique of collecting pelagic eggs, hatching them, and rearing the larvae to identifi- able size appears to be the most prom- ising approach to solving the important problem of whether bluefm tuna actu- ally spawn there. The failure to find identifiable eggs or early stages of blue- fin in the Alboran Sea (westernmost Mediterranean) as well as in the Ibero- Moroccan Bay during the 1923 and 1972 surveys, however, does not sup- port the theory of passive drif^ of spawn from the latter area to the former. 3. Western North Atlantic a. General Information Important information on the spawning of Thiomus thynmts thynnus in the western North Atlantic has been obtained through collections of larvae and juveniles. Ripe or nearly ripe fish have been captured over extensive ad- ditional areas. The great majorit\ of the captures of larvae and juveniles have occurred in the Gulf of Mexico and the Straits of Florida. Scattered occurrences farther north include one larva east of northern Florida and seven juveniles (17. 5-33. 2 mm SL) off the Carolinas, Maryland and New Jersey (Figure 67). Occurrences of ripe or nearly ripe fish have spread over some additional Table 23. Variations in the mean gonado-somatic indices'" of bluefin tuna*" taken in the trap at Moulay-bou-Selham (near Kenitra. Morocco) in May and June, 1955 (Fumestin and Dardignac 1962). Sex May 1-15 May 16-31 June 1-15 Males 1.10 (n = 21) 1.25 (n = 9) 2.04 (n = 75) Females 1.52 (n = 8) 1.42 (n=18) 1.99 (n = 91) 'The gonado- ■somatic index is 100 limes the ratio of the weight of the gonads to the total weight of the fish. '■The term n is the number offish in each sample 79 areas. They include several medium- sized (32-125 kg) bluefin as well as a few larger individuals, in the area bounded by latitudes 37°30'N to 40°30'N and longitudes 66°20'W to 70°00'W, well north of nearly all of the recorded captures of larvae and small juveniles in the western North Atlantic. Nearly ripe large bluefin tuna have also been encountered in the northwestern Caribbean Sea and northeast of the Ba- hamas, where no larvae or small juve- niles have yet been found. The capture of larvae depends not only on their abundance but also on the intensity and seasons of the collecting effort. Additional planktonic surveys in the proper seasons would probably have extended the areas of known oc- currences of larvae and small juveniles considerably. b. Specific Occurrences The most numerous and wide- spread collections of larval and juve- nile bluefin tuna fi-om the western North Atlantic area were taken in the Gulf of Mexico. During extensive ichthyoplankton surveys of the Gulf by the Centro de Investigaciones Pesqueras of Cuba in April to May 1973, and in May to June 1974, numerous bluefm tuna larvae were taken over extensive areas (Juarez, M. 1 974b, Montolio and Juarez, M. 1 977) (Figure 68). The rela- tive abundance of larvae per 1 00 m^ was calculated by the formula: NT/V X 1 GO where: N is the number of larvae per tow, T is the depth of the thermocline and V is the volume of water filtered in m'. In the 1973 cruise, which covered the eastern half of the Gulf, larvae were taken in varying numbers at 13 of the 46 stations which were occupied (Juarez 1974a, 1974b). Most of the larvae were captured in the north central Gulf at 1 1 stations within the area bounded by latitudes 25°N and 30° N and longi- tudes 87°W and 9rw, Others were taken at a station at about 26°N latitude and 84°30'W longitude, and another at about 24°N latitude and 88°W longi- tude. The April to May collection data indicated relative abundances of blue- fin larvae of up to 2,000 per 100 ml Relative abundance of larvae (number per 100 m^ of sea surface) were 2,000 for one station, 649 at an- 40" LEGEND o LARVAL <13mm + JUVENILE 13-120 mm MONTH- N = NUMBER CAPTU) 30* 20- 100' Figure 67. Collection locations for larval and juvenile bluefin nina in the western North Atlantic. other, 101 to 33 1 at three stations, and 1 00 or less at eight stations. No bluefin were captured at the other 33 stations. The estimated number of larval T. thynnus thynnus in the area, on the basis of Juarez's (1974b) data, was 3 1,442 X lO^Richards 1976). Bluefm tuna constituted 24.6 percent of all of the larvae of the family Scombridae collected during this cruise (Juarez 1974b). Bluefin tuna larvae were much more available during the May to June 1974, cruise, which sampled nearly all of the deep (over 200 m) waters of the Gulf Larvae of this species were col- lected at 23 of the 61 stations occupied (Montolio and Judrez 1977). All of the successful stations were between lati- tudes 23° and 28°30'N, and longitudes 84°30' and 94° 30'W. Nearly all of them were between 40 and 140 nauti- cal miles (74 and 140 km) from the 200 m contour. Five stations yielded more than 1,000 bluefin per 100 m^ each. These stations were rather scattered, with one at 24°N, 87°W, another at 28°N, 87°W, one at 27°30'N, 90°W, and two near 26°N, 94°30'W. From 501 to 1,000 per 100 m' r thynnus thynnus larvae were taken at five sta- tions well spread over the area. Ten stations produced 1 1 -500 per 1 00 ml and only two of the successflil stations yielded 1 00 or less. Results at one posi- tive station were not reported. No blue- fin larvae were collected at the 38 re- maining stations. The larvae of 7". thynnus thynnus constituted 47.7 per- cent of all of the larvae of the family Scombridae which were collected dur- ing this cruise (Montolio and Jaurez 1977). The May to June 1974 cruise showed that the spawning of bluefin extended over much of the deep water of the Gulf of Mexico north of 23°N latitude in that period. Comparisons with the results of the April to May 1973 cruise suggest that bluefin tuna larvae are much more available in the Gulf in May and June than in April and May, and consequently that spawning is more intense in June than in April. This is in agreement with other avail- able indications, but this comparison might have been biased by better spawn- ing success in one year than the other. "Numerous" bluefin tuna larvae 4.2-10.2 mm SL were collected at 80 23°50'N latitude, 88°40'W longitude on May 17, 1972 (Judiez 1972). Considerable quantities of these larvae were also collected near this location in the subsequent surveys of April to May 1973 and May to June 1974. T.C. Potthoff (personal communi- cation) reported the collection of one larval T. thynnus tliyrmns 5.5 mm SL June 29, 1969, and four larvae 3.6-4.2 mm SL on July 6, 1 969, in the northern Gulf of Mexico near 29°N latitude and 87°W longitude. Thus, these were cap- tured near the northeastern boundary of the area where Juarez (1974) and Montolio and Ju^ez (1977) took most of their specimens, but the dates of capture extend the time of occurrence of the bluefin tuna larvae in the Gulf of Mexico into July. Richards (1977) provided addi- tional data on occurrences of bluefin tuna larvae in the Gulf of Mexico. Nine- teen larvae ranging from 3.9 to 7.4 mm SL were collected near 27°N latitude, 96° W longitude, on May 6, 1975, and May 30- June 6, 1976. Another important series of col- lections of bluefin larvae was made off Miami, Florida, witli surface tows (fi-om 1 969 to 1 97 1 ) and on a transect from Miami to Bimini, Bahamas, with sur- face tows and oblique tows to 200 m depth (in 1975) (Richards 1976). The cumulative numbers of surface tows and captures of larval bluefin tuna are listed by half-month periods in Table 24. In total, 93 tows were made, and 1 64 identifiable bluefin tuna larvae of 3.3 to 7.2 mm SL were collected, an average of 1 .8 larvae per tow. The tows extended over a period from April 2 to July 8, but bluefin larvae were taken between April 22 and June 26 only. The number of larvae per tow was only 0. 1 in the second haH of April, but rose to 1.8 to 3.5 in the half month periods of May and June, with maxima in tlie first half of May and the last half of June. During the transects in 1975, 39 larvae 3.4-7.3 mm SL were collected in 61 surface tows at five stations, and 23 were collected in 47 oblique tows at 35" - 30" - 25" - 20" - 15 r ■ I ■■ 1 1 . ■!* LEGEND . ..;;Jt, »" - O NO LARVAE '/^f ? N/10M2 not ESTIMATED ' -.^^ .'' -»- 1-9/10M2 ■ :::j y • 10-49/10 M^ . '^x'' A 50-99/1 0M2 .:/ tf ▼ 100+/10M2 .v^jgV-?^.-..- ■; • .1 y o ■ ;^3?o.Q-+,AA+-^ •AV«« ? 4 \', . :m,^ + -»■ + AT A A + o • A 7 \'\' ''^~v, ..J ;++TAT»T •▼ + A»U ^Sj^\'^^-^ >" - •.•J^+' + ▼© + o7o •A«AoAp A^« \ ^v "^^ ^ •I ,6 7 #000 p-tj- ▼+ O O •"'• ?':?, \ V> ^ ■■^ lo + • o o o ,''o ?'-7 o + *j3^r*^N^. "■\o O O O O ^''' »? ''° <^3^!*'^7^'*^^^!^^^^3t'--'' ■/Vvooool oefJ^T^]o '""■'' ~T\^^ o - •%, oo,/' /'•.v| ' * ^^-feii ', * ' ■ " . 'If '^^^ ;o 100* 90' 80« Figure 68. Bluefin tuna larvae sampled in the Gulf of Mexico. four stations (oblique tows at station number 1 near the Miami harbor en- trance buoy were omitted because the depth of the water was insufficient). The catch rates in numbers of larvae per tow (both types combined) were maximum, 0.8 and 0.6, at stations 2 and 3. Lower catch rates, 0.1 and 0.2, were obtained at stations 1 and 4, and no larvae at all were caught at station 5 on the Bahamas side of the Straits. This distribution is rather surprising, as large bluefin tuna are observed and caught in considerable numbers along the edge of the Great Bahama Bank near Bimini in May and June but are a rarity on the Florida side (Rivas 1954, Mather 1963a). A single larva (8 mm SL) col- lected off St. Augustine, northeastern Florida, (Figure 67) in May 1 968 (T.C. Potthoff, personal communication) completes our knowledge of the distri- bution of bluefin tuna larvae in the western North Atlantic system. The known range of small juve- nile (12-120 mm SL) bluefin tuna in- cludes much of that covered by the larvae, but extends considerably far- ther north (Figure 67). Rivas (1954) collected a 45 mm SL juvenile off Mi- ami on June 9, 1953. Potthoff and Richards (1970) identified 40 juvenile bluefin 22-118 mm long which had been regurgitated by tems during a tem- banding operation at the Dry Tortugas, Florida, in June and the first half of July 1960-1967. None were found in May, although juveniles of other scombrids were collected. No scombrids less than 20 mm long were collected in this manner. Twenty addi- tional juvenile bluefin in the same gen- eral size range and season have been obtained from tems at the Dry Tortugas subsequent to the 1970 publication (T.C. Potthoff, personal communica- tion). T.C. Potthoff (personal commu- nication) also informed us of five juve- nile bluefin 22-33 mm long collected in the Gulf of Mexico in late May and early June, and four others, 16-21 mm long collected off the Carolinas in May. A most surprising record, also provided by Potthoff was of a 17.1 mm SL juvenile collected off Mobile, Alabama, on August 19, 1954. This specimen must have been spawned long after adult bluefin have usually left the Gulf, and also in a month when the extensive Cuban ichthyoplankton survey in 1973 failed to collect a single larva of this species (Judrez 1974b). Three juveniles 21-33 mm long collected off Maryland (37°42'N latitude, 73°10'W longitude) and New Jersey (38°45'N latitude, 71''00'W longitude) July 27-28, 1959, have been identified as bluefin tuna by Watson (Watson and Mather 1 96 1 ) and this identification has been verified (T.C. Potthoff, personal communica- tion). These records are the most north- erly in the western Atlantic for small juvenile bluefin. They are of special interest because they are from the area where medium-sized bluefin may spawn (Matiier l974,Baglin 1976). The spawning habits of these fish in the western Atlantic are little known, but they do not occur in numbers anywhere near the known spawning areas of the large bluefin (Madier 1963a; also see Section IV). Examinations of gonads suggest that the spawning grounds of the blue- fin tuna in the western North Atlantic may be much more extensive than is indicated by the distribution of catches of larvae, or even of juveniles (Mather 1974, Baglin 1976). Rivas ( 1 954) examined the gonads of 95 large ( 1 99.7-255.0 cm long) blue- fin tuna caught near Bimini, Bahamas, all in the month of May, but in three different years, 1952, 1953 and 1954. He classified the specimens, by the color and consistency of the gonads, the na- ture of the ovarian eggs, and the amount of milt present, as recently spent, partly spent, or ripe. All of the 29 males examined were classed as recently spent. He also con- sidered that 6 1 of the 66 females exam- ined were recently spent, but one was j udged to be partly spent, and four were considered ripe. The smaller eggs of ripe females were non-spherical, opaque and mea- sured 0.4 to 0.7 mm in diameter, with a mode at about 0.6 mm. The larger eggs were spherical, translucent, yolk-filled, and measured 0.7 to 1.1 mm in diam- eter, with a mode at about 0.9 mm. Rivas ( 1 954) offered two possible ex- planations for the small number of ripe individuals (about 4%) found in com- parison with the larger number (about 96%) of spent individuals. Since the fish which are observed in the area are nearly always travelling northward, it was possible that they had spawned in an area south, or southwest, of the fish- ing area near Bimini and were caught when spawning had been nearly com- pleted. He assumed that they had mi- grated to this area through the Straits of Florida, passing close to Havana, Cuba, and the western edge of Cay Sal Bank. The other possible explanation which Rivas offered was that the rod and reel fishing method used at Bimini selected spent rather than ripe fish. He quoted previous authors on the reluctance of spawning fish to feed. During cruise 57-5 of the Bureau of Commercial Fisheries M/V "Dela- ware", R. H. Gibbs and B. B. Collette, then of Woods Hole Oceanographic Institution, macroscopically examined the gonads of 48 bluefin tuna caught during the period June 8-14, 1957, in the area between latitudes 37°30'N and 40°30'N and longitudes 66°W and 70°W. Their observations were repro- duced by Baglin (1976). Two age-3 (95-105 cm) individuals were consid- ered immature. The testes of one male of age 4(121 cm) contained abundant milt. Two other males and two females of this age group were classed as im- mature, but the ovaries of another age- 4 female were "much vasculated," a sign of ripening. Milt was squeezed from the testes of a 5-year-oid (140 cm) male. The ovaries of a female of the same age-size group were preserved for laboratory examination, presumably because of signs of maturity. Six males and six females, probably 6 years old ( 143.5-1 57.2 cm), were examined. The testes of all of the males contained abun- dant sperm; those of one individual were much enlarged, and another was classified as ripe. One female was classed as "near ripe." The ovaries of another contained eggs which seemed to be nearly ripe, but were not loose. Those of two others were well devel- oped and loose, indicating imminent spawning. The ovaries of the last two females of this group contained no eggs, and were believed to be spent. These observations, and less nu- merous ones from later catches of the Table 24. Number of surface tows and number of larvae collected 25 nautical miles (4.0 km) from the Miami harbor entrance (Station 1) (1969-1971 and 1975) and additional stations (2-5) completing a n-ansect of the Straits of Florida from Miami to Bimini, Bahamas, in 1975'. Data are from Richards (1976). Station 1 (1969- 71, 1975) Stations 2-5 (1975) Total Date Tows Larvae Larv. /Tow Tows Larvae Larv./Tow Tows Larvae Larv ./Tow April 1-15 6 8 14 April 16-30 5 1 0.2 4 9 1 0.1 May 1-15 8 45 6.7 8 7 0.9 16 52 3.2 May 16-31 6 6 1.0 8 20 2.5 14 26 1.8 June 1-15 8 23 2.9 8 10 1.2 15 33 2.0 June 16-30 10 52 5.2 5 15 52 3.5 July 1-12 4 5 9 Total 47 127 2.7 46 37 0.8 93 164 1.8 ♦The vessel and nets used in 1 969- 1 97 1 differed considerably from those used in 1975 82 "Delaware" in this area and season, indicate that at least some of the me- dium-sized and the larger "small" blue- fin have spawned in May and June in the waters north of the Gulf Stream off the northeastern United States. We have noted (Section VC 1 ) that ripe or nearly ripe bluefin may have travelled consid- erable distances in short periods before their capture. In this case, however, Wathne (1959), Wilson and Bartlett (1967), and all the additional data we have been able to collect on bluefin tuna catches in the western North At- lantic, indicate that medium-sized blue- fin have occurred north of the Gulf Stream during the winter and spring but have been only rarely captured south of it and west of 60^W longitude. It therefore seems unlikely that these fish had migrated from southerly areas im- mediately before their capture. Baglin (1976) presented a histo- gram showing the "gonadal-somatic in- dex" (the "gonosomatic relation" of Rodriguez-Roda 1 964a) by months for 67 bluefin tuna weighing over 100 kg (Figure 69). The specimens were from various areas, but nearly all of those captured in May and June, when the index was at its maximum, were from the vicinity of Bimini and Cat Cay in the Bahamas. The mean gonadal-so- matic indices were about 1.5 in May and 1.0 in June, as against values of from 0.25 to 0.5 in other months. The sex ratio for 237 large bluefin tuna which had been captured near the Bahamas in the years 1950-1966 and examined by Woods Hole Oceano- graphic Institution and National Ma- rine Fisheries Service personnel (un- published data) and Rivas (1954) was 72% females and 28% males. Baglin ( 1 976) also showed the fre- quency distributions of ovum diam- eters for bluefin taken at various dates and in various conditions. He found no developing eggs in bluefin tuna less than 10 years old. He set up a maturity scale based on the gonadal-somatic in- dex, the diameter and morphology of the ova, and the appearance and physi- cal properties of the gonads. He con- cluded that the bluefin spawned in May and June, and probably in April also. He observed that spawning must start south of Bimini and, on the basis of the data from "Delaware" cruise 57-5, might extend as far north as off New England, and might involve smaller fish. Characteristics of the gonads of giant bluefin tuna collected during ex- ploratory fishing cruises of United States and Russian research vessels in- dicate that these fish may spawn in the northwestern Caribbean, the Windward Passage, the old Bahama and Santaren Channels, and a large area east and north of the Bahamas (Figure 70). Their presence in numbers in the latter area during the spawning season is con- firmed by catches of the Japanese longline fishery (Fisheries Agency of Japan 1971, Wise and Davis 1973). The collection data and sizes of the fish, and their gonad weights and ma- turity indices are shown in Table 25. Zharov ( 1 965) observed that large bluefin taken by longline in late May and early June 1963 north of the Baha- mas and east of central Florida (near 29°00'N, 79°00'W, and 30°00'N, 77°30'W) were "typically spawning, since their sexual products were in stages IV to VI-II" (these stages were not defined). The fish were from 1 98 to 238 cm (average 219.4 cm) long, and weighed from 140 to 220 kg (average 177.5 kg). The information summarized above defines some positive locations and periods of spawning of the bluefin tuna, and more extensive tentative ones. The best documented spawning area is in the Gulf of Mexico (Figure 68). Bluefin larvae have been collected over much of the deep (more than 200 m) area of the Gulf north of 25°N, and also off the northern edge of the Campeche Bank, near 23°30'N and 94°30'W, and at 23" 30'N, 85°W. The most thorough and extensive surveys of the area were carried out in April- May 1973, and May-June 1974 (Juarez 1974b, Montolio and Juarez 1977). A few specimens have been collected in early July, and one, reportedly, in Au- gust. The last specimen, a 17.1 mm SL juvenile collected in the northern Gulf in August, almost certainly represented an aberrant occurrence, or a case of incorrect collection data, since the northward migration ofspent adult blue- fin through the Straits of Florida ceases before the end of June (Rivas 1954, Mather 1 963a), and no adults of the species have been recorded in the Gulf of Mexico in late July or August, to our knowledge. These same considerations, and the fact that the authors did not list the characters on which they based their larval identification, also cast doubt on Gorbuno va and Salabarria's ( 1 967) re- ports of bluefin tuna larvae in the Gulf 5'^ Q 14 ^ 1.0 §0.8 ^06 §04 §0.2 JAN MAR APR MAY JUNE JULY AUG SEPT OCT MONTH Figure 69. Seasonal variation in mean gonadal-somatic indices of western Atlantic bluefin tuna greater than 100 kg (number offish indicated above bars) (Baglin 1976). 83 of Mexico in September, and in Cuban waters in September and October (Potthoffand Richards 1970, Richards 1976). The major spawning of giant bluefln in the Gulf of Mexico most probably takes place in May and June, with lesser occurrences in late April and early July. The distribution of catches of lar- val bluefin tuna along the transect of the Florida Straits from Miami to Bimini, with the largest catches at the western and central stations, and none in the station off Bimini, seems very significant. It suggests that the majority of the larvae have been transported by the Florida current from areas in the southwestern part of the Straits, and possibly even in the southwestern Gulf of Mexico. Conversely, it does not in- dicate extensive spawning in the area of the fishery for adult bluefin along the northwestern edge of the Great Bahama Bank. This in turn strongly supports the belief that the majority of the large adult bluefin which pass Bimini in May and June have spawned before arriving there, as did the studies of their gonads (Rivas 1954, Baglin 1976). Rivas' (1954) alternative sug- gestion that the northward migration might include schools of spent fish, which may take bait, and schools of ripe fish, which do not, seems less prob- able in view of this new information on the distribution of larvae. It thus ap- pears that this migratory passage off the northwestern Bahamas is actually a spent fish run, analogous to the "re- turn" runs which occurred regularly along the north coast of the Ibero-Mo- roccan Bay and in some Mediterranean localities. Rivas (1954), Potthoffand Richards (1970), Richards (1976) and Baglin (1976), on the basis of their studies of gonads, larvae and juveniles collected in or near the Straits of Florida and the occurrences of spent adults along its eastem edge, reached slightly different conclusions about the prob- able dates of spawning in that area. All of their estimates, however, fell within the months of April, May and June. The total evidence suggests that some spawning may have occurred in late April, but that the bulk of it has taken place in May and June. The termina- tion of spawning may well have coin- cided with the disappearance of the adult fish, which has usually occurred during the last half of June. As noted above, the collections of bluefin larvae in the central and west- em parts of the Straits of Florida off Miami suggest that spawning may have occurred in the Straits south and west of there, off the Florida Keys or possi- bly in the southwestern Gulf of Mexico. Bluefin tuna have very rarely been ob- served off the Florida Keys (Section IV), but, as Ju^ez ( 1 974b) pointed out, the species tends to spawn in offshore waters. Most of the very large sport fishing effort exerted off the Florida Keys has taken place on or close to the continental shelf This might explain the scarcity of observations of bluefin, even if they had been spawning in the deep waters off the Keys. E. RELATIONSHIPS BETWEEN ENVIRONMENTAL FACTORS AND SPAWNING BEHAVIOR 1. Introduction Biologists agree that the spawning behavior of bluefin is strongly affected by environmental factors and that its sensitivity to these factors is intensified during the reproductive period. Opin- ions as to the importance and the nature of the effects of various conditions, however, are extremely diversified. Among the environmental factors most frequently considered are the tem- perature, salinity, density, transparency and oxygen content of the water, winds, tides, currents, atmospheric pressure, rainfall, and abundance of plankton. Most of this research has centered on mature individuals. Less attention has been paid to the effects of environmen- tal conditions on the hatching of eggs and the survival of the early stages. With the present decrease in the spawn- ing stocks, however, knowledge of these factors may also be very important. 2. Mediterranean Sea Since the ancient trap fisheries in the Mediterranean Sea depended on the spawning behavior of the bluefin tuna, much attention has been paid there to the effects of environmental factors on this behavior. Roule(1914a, 1914b, 19I7),apio- ncer in these studies, concluded that the bluefin tuna was "stenothermic" and "stenohaiine." These beliefs were the basis of his "halo-thermic" theory. Roule did not specify the limits, or averages, of the temperatures and sa- linities which were suitable for the blue- fin. He did, however, note that the sen- 60' BERMUDA t2:> Figure 70. Collection locations for the large, mature bluefin tuna collected during exploratory fishing cruises of United States and Russian research vessels (see Table 25). 84 ■s o 00 ■a ?> V) a. E '^ u w u « a o ±: O at> O -3 ^ E u 2 § « S B O _ « U ^ i C/3 Z Ti-\ot-;r-TrTrr^fnf<^\o\q^^oooooooooo — _______ — — ' — — — tN(NrN(NtN(NfN(N(N(Nr^r~l(Nr^(-Nlr)(N(N(>4(N (Nr^l(Nri_-rnrnO00vpTrfN0000fNO vor^oO"^ooOt^m — Ov lo O — ; fN T fN r/^ ^ 00 o r^ __ O m •^ r^ r~ o ir> vo On __ o o ■/-l 00 rvi w-i >/-l in o o (N ^, ^^ C- C- C- v_^ C- (N N_-' ^— ' •^^ c C- C- C- (^1 On— lOviJ-i^v-iw-iini^oomvO f-iOOvOr^inv^OOO- — r^"/ ■ — mocQNrnqp^— -■^riooONON _r'\-^ 3 M UU .S ^^^^^^^^^^^^^^^^^^^^^^^^^^3:^^^^ O m I-~ r^ r^ I^ r-~ ■* r^ [^ r^ o r~ o NO NO oo oo 00 oo Tt TT o o o So So So n TT "* rr ■* ■^ ■* r<-i •■■ ■~ o •^ CN) r~4 o O o o rn rn rl r'-i m r4 r-i rN4 o o o o ^ o '3- o o o o o o 00 ^ o o h ^ o o o «n o »n o o •n v°n o NO v°rv o NO o NO oo o 00 00 On o On o fl r- r-- [^ r^ r~- r~ r~ NO VO r^ t^ r^ r- r~ 1^ r^ t^ r^ r^ t^ (^ 1^ t^ r^ r- r^ r- r~ r~ r^ f- ^ S Z Z z Z Z Z Z Z Z z Z Z Z Z z Z z Z z Z Z z z Z z Z Z z z z z z m U-l — _. — :— — < ON •^ oo oo rN4 oo VO >n v-i o o o o fs ?N| CNl rs r-4 ON ON On f~> <-> o E 3 C > > > > > > > > > > > ^ > > > > > > > > > > > > > > > > > > > - (N m ■* 'T 'I- ■<«• "* »n NO i^ r- 00 o- o - - rsi (N (^1 CM r^ r^ m rn en ■* ^ ■^ ' impor- tant factor, as was demonstrated by the catches of two traps near the mouth of the Guadiana River at the Portuguese border. About 90 percent of the fish were caught in clear water. The re- mainder were taken in "regular" or "dirty" water, without a consistent dif- ference between the catches in these two types. The clarity of the water, however, seemed more critical during the "return" (westward) movement than during the "arrival" (eastward) transit. It was difficult to find a definite rela- tionship between winds, or wind-driven currents, and the catches of the traps. De Buen concluded that the direction, intensity and frequency of the winds might act as a secondarj' factor influ- encing the local movements of the tuna by altering the temperature of the sur- face waters. In discussing the results, F. de Buen (1925) stated that the bluefin tuna was a typically stenothermic fish, whose disffibution was exactly limited by tem- perature. In the spawning season, its sensitivit) to temperature increased. In his opinion, the tuna was a surface- dwelling fish, submerging occasionally to depths of only a few meters. He rejected the abyssal winter habitat pro- posed by Pavesi ( 1 887) and supported by Sanzo ( 1 910a). He believed that the adult tuna which inhabited the western Mediterranean basin were altogether independent of those occupying the adjacent parts of the Atlantic. He pos- tulated that the colder waters of the Strait of Gibraltar separated the two basins oceanographically and prevented the interniinghng of these two groups of tuna. De Buen concluded, from his own observations, that the water tempera- ture was the main factor infiuencing the movements of the tuna, and pro- posed a thermic theory, as opposed to the hydrodynamic theory of Bounhiol (191 1) and the halothermic theory of Roule (1914a). F. de Buen (1925) found, by com- paring the Spanish trap catches of May 1923 with oceanographic observations made concurrently in the same area, that the maturing "arrival" tuna pre- ferred the warmer and less saline wa- ters near Cadiz to the adjacent colder and more saline waters, which are typi- cal of the open ocean to the west and the Strait of Gibraltar to the east. After the peak of the "arrival" fish- er) along the coast, the fisher> dimin- ished. This coincided with the greatest intensity of spawning In general, the conditions causing variations in the wa- ter temperature, such as the persistence of certain winds and the influence of coastal currents, were secondary fac- tors atTecting the local movements of tuna. Fresh waters carrying suspended materials might have caused the tuna to leave the coast and temporarily pre- vented their capture by certain traps. The spawning areas coincided with the centers of high temperature, varying constantly under the effect of coastal factors on one side, and the high seas on the other. F. de Buen stated that affer spawn- ing, the tuna, no longer subject to such 88 close constraints in regard to tempera- ture, dispersed in the Ibero-Moroccan Bay, contributing another period of great catches for the traps (the return). The maximum catches in the "return" fishery, however, were concentrated at the western and eastern ends of the southern Spanish Atlantic coast, at or near areas where the temperature and salinity were considered the least fa- vorable during the "arrival" period. O. de Buen (1924), in a prelimi- nary report on this survey, presented the following conclusions: The moment of spawning changed from year to year according to changes in environment. Increases and decreases in water temperature exercised a great influence on the development of the gonads. High temperature, high salinity and agitated waters with high dissolved oxy- gen content were sought by the tuna for spawning. The temperature of the surface wa- ter in the Strait of Gibraltar was 3°C lower than that of the coast of Cadiz, and 5°C less than that of the Mediterra- nean coast of Morocco from Ceuta east- ward. These conditions were unfavor- able for spawning tuna, and constituted an obstruction to their migration through the Strait. On the other hand, the currents and winds in the Strait increased the dissolved oxygen in its waters, a favorable factor for such mi- grations. The tuna spawned west of G ibraltar before reaching Tarifa, where the sa- linity became lower. The next important survey of the environment in relation to the occur- rences of spawning bluefin tuna in Spanish waters was by Lozano Cabo (1957, 1958). He studied oceanographic conditions at the Barbate trap east of Cadiz and the biology of its catches during the 1954 fishing season. He made similar studies in the Moroccan trap fishery (Section VE3d) in 1954, providing the basis for interesting com- parisons of data from the two areas. The surface temperature at Barbate in June and July varied from 1 7°C to 2 1 °C ( 1 8.9° +/.0.089°), increasing pro- gressively despite small local alter- ations. There were greater fluctuations, from 14°Cto 19°C (16.83° +/-0. 146°), in the temperature at a depth of 35 m, which decreased, instead of increasing as did the surface temperature, in the final days. The water was colder at 35 m than at the surface. Lozano Cabo found a negative cor- relation index of -0.29 with a probable error of +/-0. 1 1 between the surface temperature and the catch. There was no correlation between the temperature at 35 m and the catch. The "arrival" tuna preferred water temperatures from 18°C to 21.5°C off southern Spain and Morocco, with a more specific prefer- ence for 18.2°C to 18.7°C at Barbate. The transparency of the water, measured by Secchi disc, varied from 5 m to 17 m, with a mean of 13.04 m, at Barbate. There was a correlation index of 0.40, with a probable error of +/- 0. 1 0, between the transparency and the catch. The best catches occurred at transparencies of more than 13 m and especially over 15 m. Minimal catches always occurred at transparencies of less than 14 m. The fishermen believed firmly that the catches of the traps depended inti- mately on the lunar phases and the tides. Lozano Cabo could not confirm this, except that, when tuna were present, they tended to enter the trap with rising tides. Tliis might signify a negative rheotaxis, at least during their pre-spawning migration. The correla- tion coefficient of +0.10 with a prob- Table 26. Surface temperatures and salinities at the trap off Cabo de Santa Maria and catches of the traps in the Algarve (southern Portugal) in the period from late April through August by month. TEMPERATURE AIVD SALINITY OF THE SURFACE WATER OFF CABO DE SANTA MARIA Month Temperature (°C) Min. Max. Avg. Salinity (VJ Min. Max. Avg. April" 15.0 19.5 17.4 35.93 36.02 35.97 May 15.0 20.5 17.3 35.77 36.15 35.99 June 15.0 20.0 17.4 35.77 36.22 35.96 July 15.5 20.0 18.7 35.82 36.11 35.96 August 17.0 26.0 22.3 35.73 36.29 36.04 FISH" RECEIVED BY THE CANNERIES. 1958 "Atuns" "Atuarros" 'Albacoras""Cachorretas" I'otal May Weight kg 73,166 8,713 No. offish 439 119 Average kg 166.7 73.2 June Weight kg 307,071 130,658 No. offish 2.330 1,903 Average kg 131.8 68.7 July Weight kg 65,870 4,746 No. offish 457 64 Average kg 144.1 74,2 August Weight kg 21,282 1149 No. offish 157 15 Average kg 135.6 76.6 •April 24 - 30 •"Size groups of tlsh (Vilela 1960): Atuns '- Albacoras = 30-49 kg, Cachorretas ' 30 kg. 723 115 82,717 12 7 577 60.3 16.4 143.4 3,052 500 441,291 68 78 4,379 44.9 6.4 100.8 244 - 70,860 5 - 526 48.8 - 134.7 932 - 23,363 17 - 189 54.8 _ 123.6 90 kg, Atuarros = 50-89 kg. 89 able error of +/-0. 12 between the state of the tide and the catch of Barbate was not significant. Salinities and dissolved oxygen were also observed at Barbate by Lozano Cabo. Although measurements of both were taken from the surface and at depths of 30 m or 35 m, he did not study the correlation with the catches. He had found no correlation between salinity and catch at the Los Cenizosos trap in Morocco. Salinities at Barbate fluctuated between 35.39 o/ GO and 35.53 o/oo at 35 m. Dissolved oxygen levels were abnormally high varying from 6.9 to 8. 1 cc/l at the sur- face and 5.0 to 9.0 cc/l at 30 m. Rodriguez- Roda (1963, 1965, 1969c, 1970a, 1970b) collected envi- ronmental data at the Barbate trap from 1961 to 1969, and discussed their pos- sible relationships to its catches and the movements of the maturing and post- spawning bluefm tuna. His 1963 publi- cation presented monthly values of the temperatures, phosphate production, and salinity, of the water at various depths from the surface to 30 m, and its transparency and optical absorption. Rodriguez-Roda (1969c) pre- sented the mean temperatures at 1 m intervals from the surface to 30 m in the Barbate trap in August 1967. Rodriguez-Roda ( 1 970a) collected daily temperature measurements of the wa- ter temperature at 10 m intervals from the surface to 30 m for the months of May through August 1968. He com- pared the weekly averages of these tem- peratures graphically with the corre- sponding numbers of tuna caught in the trap. He (1970b) collected corre- sponding data and presented a similar graphical comparison for the same pe- riod in 1969. In that year he also mea- sured the transparency of the water each day, and he compared these values graphically with the daily catch of the trap in numbers of tuna. The temperature during the most productive months. May to July in 1 96 1 and 1962, ranged from I8.0°C to 2 1 .4°C at the surface and 1 6.5°C to 1 8° C at a depth of 30 m. During the "ar- rival" run in May and June 1968, the highest catches occurred with mean surface temperatures of I7.2°C to 18.9°C and temperatures of I6.6°C to 17.6°C at 30 m. The best catches dur- ing the 1 969 arrival run were taken in a week with mean temperatures of about 1 8°C at the surface and 1 6°C at 30 m. During the "return" period in July and August 1 968, the best catches took place with temperatures of I9.5°C to 19.9°C at the surface and I6.3°C at 30 m. In 1969, the best "return" catches were taken with mean surface temperatures of 21.0°C to 22.5°C, and temperatures of 1 8.5°C to 20.2°C at 30 m. Rodriguez- Roda noted (1969a), that the August temperatures in 1961, 1962 and 1967 were about the same, and that the total production of the Spanish Atlantic traps in those years did not vary greatly. In his 1 970b publication, he obser\ ed that the catch of these traps was greater in 1 969, when the water was wanner, than in 1968. Rodriguez-Roda (1971) con- cluded that the maturing "arrival" pe- riod begins at temperatures of 16°C to 1 7°C, and is most productive at tem- peratures of 18°C in the upper 10 meters, 17°Cat20mand 16°Cat30m. In the post-spawning "return" period, optimum temperatures are 20°C to 2 1 °C in the upper 10 m and I9°C at 30 m. The average salinities during the monthsofMay, June and July in 1961 and 1962 varied from 35.88 o/oo to 36.34 o/oo at the surface and 36.01 o/ 00 to 36.21 o/oo at 30 m (Rodriguez- Roda 1963). The transparency of the water ranged from 1 1 m to 2 1 m in May and June of 1961 and 1962, and from 2 m to 25 m in May-August 1969, (Rodriguez-Roda 1963, 1970b). The author found a high positive correla- tion, during intensive periods of the fishery, between the transparency and the catches of the trap. Phosphate production in May, June and July, the months of ma.ximum tuna catches, varied from 0.54 to 0.59 mg/l at the surface. Chlorophyll production was gen- erally greater in the cold months than the warm ones, with a large maximum in October and lesser ones in January' and April. Zooplankton was predominant in the wami months, although undergo- ing many variations during the year, due to the influences of Atlantic wa- ters. Rodriguez-Roda ( 1 963) concluded that the Barbate area, in regard to its oceanography and the plankton in gen- eral, could be said to be under the influ- ence of Atlantic waters, especially in its surface layers. d. Morocco Lozano Cabo (1957, 1958, 1970) conducted thorough studies of the en- vironment at the Los Cenizosos trap, north of Larache, Morocco. Our ac- count generally follows his 1970 sum- mary, with some details added from the earlier reports. The water temperature varied from 1 7°C to 22°C at the surface, from 1 5°C to 20°C at 15 m, and from 14°C to 1 8°C at 35 m. The correlation between the surface temperature and the num- bers of flsh caught was negative (-0.72 +/-0.06). Fish were caught at tempera- tures between 17°C and 21°C, but mainly between l8°Cand 19.7°C. The transparency of the water ranged from 8 m to 19 m, with an average value of 13.7 m. A high posi- tive correlation, + 0.59 +/-0.08, existed between the transparency and the num- bers of flsh caught in the trap. The catches were poor when the franspar- ency was less than 13 m and good when it was over 1 5 ni, especially at 1 5 m to 16 m. The salinities were between 36.00 o/oo and 36.34 o/oo at the surface, 36. 1 1 o/oo and 36.33 o/oo at 1 5 m, and 36.06 o/oo and 36 25 o/oo at 35 m. The correlation between surface salinity and the numbers of tuna caught in the trap.- 0.078 +/-0.13, was not significant. Likewise, no correlation was found between the height of the tide (above the daily low water level) and the num- ber of tuna caught, even though the captains thought that the tides strongly influenced the catches. There was a clear relationship, however, between the velocity of the tidal currents and the presence of tuna. Good catches coin- cided with tides of large amplitude, during which the coastal currents were stronger and farther from shore. The dissolved oxygen content var- ied from 5.9 to 7.9 cc/l at the surface, from 5.7 to 7.9 cc/l at 15 m, and from 5.5 to 7.7 cc/l at 30 m. Alone le ( 1 964. 1 969) developed a hypothesis on the migrations and spawning of bluefln tuna, and the in- fluences of water temperature upon 90 them, in the waters between the south- em Iberian coasts and the Canary Is- lands. This hypothesis was derived froni studies of the tunas and their fisheries off the Atlantic coast of Morocco, and a hydrographic survey of the above waters. In early June 1964, tuna were found between Lanzerote Island (Ca- naries) and the African coast in water with a temperature of 20.06°C and sa- linity of 36.40 o/oo at the surface, and 18.37°C and 36.56 o/oo at a depth of 50 m. Aloncle believed that these tuna were pushed northward with the sea- sonal warming of the water, and kept away from the Moroccan coast by iso- therms of temperatures greater than 2I°C. In the course of this movement, he believed that they spawned in the area between Lanzerote, Conception Bank and the Moroccan coast. e. Other Eastern Atlantic Areas The only other parts of the eastern Atlantic where bluefin tuna are believed to spawn are in the vicinit>' of the Azores (Ferreira 1932) and in equatorial wa- ters south of Sierra Leone (Richards 1969, Richards and Simmons 1971, unpublished data reports for Geronimo cruises 3, 4 and 5). In May and June, the months in which bluefin reportedly spawn around the Azores, the surface temperature in the area varies between about 1 7.0°C and about I9.7°C, while the salinity remains at about 36. 1 o/oo. In the area bounded by latitudes 7°N and 8°S, and longitudes 1 3°W and 15°W, three larval T. ihynnus thynmis were collected in March 1963, in wa- ters with surface temperatures exceed- ing 27°C and surface salinities of from 36.4 o/oo to 38.8 o/oo (Richards 1 969). These temperatures are much higher than those in other reported spawning areas of the species in the eastern At- lantic and the Mediterranean. 4. Western Atlantic a. Introduction In comparison with what is avail- able for the eastern Atlantic and the Mediterranean, ver>' little has been pub- lished on the relationships between the spawning of bluefin tuna in the western Atlantic and environmental conditions. Therefore our discussion of this corre- lation depends almost entirely on ob- servations which weie made concur- rently with investigations, but have not been synthesized or analyzed, and data in atlases or survey reports. b. Gulf of Mexico The best documented spawning of bluefin in the western Atlantic takes place in the Gulf of Mexico in from late April to early July (Section VD3). Tlie heaviest concentrations of larvae evidently occur in the deep (more than 200 m) area behveen latitudes 23°N and 30°N and longitudes 84°W and 94°W (Figure 68) (Juarez 1974b, Montolio and Juarez 1977). During April, May and June the surface tem- perature of these waters increases from between 22°C and 24°C to a little over 2TC (Galtsoff 1954). The temperature in the upper layer is fairly constant down to 30 m. TTie salinity in the upper 50 m layer is typically 36.00 o/oo. In contrast, the salinity of this layer over the Campeche and Florida Banks, where few larvae have been collected, is 36.25 o/oo, possibly because of up- welling. c. Straits of Florida The Straits of Florida at the Mi- ami-Bimini line present an unusual situ- ation in that adult bluefin are found in numbers, and almost exclusively, on the eastern (Bahamas) side (Section IVC2), but the abundance of larvae is minimal on the eastern side and maxi- mum near the Florida side (Section VD3). Most of the fish examined at Bimini were spent, with a very few ripe ones among them (Rivas 1954). Evi- dently most of the spawning occurs south of this line. The surface water temperature across the Straits of Florida between Miami and the Florida Keys and the Bahamas (25°N to 26'N and 79°W to 80°W) is fairly unifomi in the deeper central area. On the edges of the Strait where the bottom rises sharply tem- peratures are cooler (Sverdrup et al. 1942). The surface temperature in the deep water is about 26°C to 26.5°C in April, increasing to about 27°C in May and about 28°C in June (Pyle 1962). ' Nearly all of the catches of bluefin tuna in the Straits of Florida have been taken by rod and reel from schools traveling close to the surface (Rivas 1978). The fish are seen near the sur- face in this area only under certain con- ditions of wind and cuiTL-nt, however, and it is presumed that the\, must spend much of the migrating period at deeper levels. A male with milt was taken by longline at an estimated deptli of 55 m on the eastern side of the Straits just north of Bimini (Mather and Bartlett 1962). The water temperature at the location was 26.7°C at the surface and 25.7°C at 55 m. The salinity in the upper 50 m layer in this region ranges from 35.5 o/oo to 37 o/oo, with an average of about 36.2 o/oo, from April through June 1977. The average phos- phate and oxygen levels in this period were about 0. 1 and 4.65 ppm, respec- tively. d. East and North of the Bahamas In the waters east of the Bahamas the research vessel Crawford caught bluefin with maturity indices from 28 to 126 (mostly 60-90) at several loca- tions in May 1961 (Figure 70, Table 25). Fhe surface water temperatures ranged from 23.6°C to 26.5°C, but most of the fish were taken in waters of 25°C. At 55 m the temperature range was 22.4°C to 25.3°C, but most of the fish were taken where the temperature was about 24°C. A little north of the Bahamas the Crawford caught eight fish with matu- rity indices of 21 to 98 (Table 25) in water with surface temperature beUveen 24.2°C and 25.6°C and temperature at 55 m between 20.8^C and 24.1 T. Zahrov (1965) reported that spawning fish caught north of the Bahamas (Fig- ure 70) at the end of May and the beginning of June were in water with surface temperature of 25°C to 26.4°C. e. Northeastern United States The bluefin showing signs of spawning taken by the M/V Delaware about 500 km off southern New Jersey and the Delmarva Peninsula during cruise 57-5 (Section VD3) were gener- ally in cooler waters than were any of the previous fish. These catches were made in waters of surface temperatures between 1 8.3X and 25.5°C. The most successful catches were in water w ith a temperature around 20' Cor 2 rC. Fish- ing depths were estimated to be 52 m to 57 m and most of the fish were taken where the water temperature at this depth ranged from 1 5 .5 "C to 2 1 . 1 °C. 91 F. Feeding Activity During the Spawning Season Different opinions have been ex- pressed in regard to the feeding activity of bluefin tuna during the spawning season. Many of the studies used indi- viduals caught in traps. It has been argued that the lack of food in the stom- achs of "arrival" fish might have been caused by digestion while the fish were awaiting removal from the trap, or by regurgitation resulting from their struggles during this removal. Food is frequently found in the stomachs of "return" fish, however, even though they undergo the same treatment. F. de Buen (1925) and Sella (1929a) were of the opinion that tlie bluefin continued to feed during their period of maturation and spawning, but at a reduced rate. De Buen believed that they ate food which they happened to encounter, and could catch without undue effort, but seldom exerted them- selves in the pursuit of difficult prey. Rodrfguez-Roda (1963, 19643, 1 969) found that stomachs of tuna taken in the Spanish "arrival" traps were usu- ally empty. Some contained a few swimming crabs, Polybins hens/owi Leach, which they had presumably con- sumed before entering the trap. Small fish which had been caught in traps with the tunas were occasionally found in their stomachs. He emphasized the limitations of his findings because of the possible losses of stomach contents of trapped fish mentioned above. Sara's (1964, 1973) studies of stomach contents of "arrival" fish taken in the fraps off western Sicily and ob- servations of their behavior when con- fined in the traps with species on which they normalK fed, led him to believe that maturing bluefin mna abstained from eating. He found increasing amounts of food in the stomachs of post-spawning fish, however, as the season advanced and their gonads be- came smaller. Arena ( 1 964) found stomachs of "arrival" fish caught in the western Si- cilian traps to be empty, or nearly so. Those of others caught off the eastern part of the island, however, contained considerable quantities of food. A large quantity of swimming crabs, Polybhis henslowi Leach, was found in the stom- ach of the first group of bluefin tuna caught atMilazzo in 1 96 1 . Several other authors have reported observations of food in the stomachs of maturing blue- fin caught off eastern Sicily (Genovese I960, Genovese and Alonzo 1961 Li Greci 1961). The considerable longline catches of bluefin tuna in the central Mediter- ranean during the spawning seasons of 1973 and 1974 (Fisheries Agency of Japan 1975, 1976) prove that signifi- cant numbers of these fish feed during the spawning season. As Sara (1973) noted, the trophic (feeding) tendency may on occasion overcome the tenden- cies which are usually dominant dur- ing the spawning period, including the tendency to reduce, and finally stop, feeding as the volume of the gonads increases. G. The Spawning Act The only published accounts of the actual spawning of bluefin tuna which we have encountered described occurrences in the central Mediter- ranean. Some of these activities oc- curred while the fish were being held in traps, but others, probably more im- portant because the fish were under no constraint, took place in the open sea. Sella (191 1) and Held! (1932) de- scribed the spawning actions of fish in traps off Sicily and Tunisia on the basis of accounts which they considered reli- able. Sara (1964) personally observed the reproductive acts of tunas in the vicinity oftrapsofifSicily. Arena (1964) likewise witnessed the spawning of bluefin tuna in the open sea off Sicily. These accounts show that spawn- ing occurs at the instant when a pair of fish turn on their sides and make con- tact, or appear to, with their ventral surfaces. This refutes the previous sup- position (Tiews 1963) that tuna spawned by saturating an area with randomly discharged eggs and milt. Arena and Sara, however, reported that small groups of tuna sometimes engaged in communal mating. In a spec- tacular observation, Arena (1964) saw files of 10-12 giant fish overtake other files of about the same number, with the fish of one file mating with those of the other as they passed. P. Arena (personal communica- tion) provided the following informa- tion on spawning, based on his more recent observations. The mating offish in large schools may involve nearly all of the fish in the school at the same time. This causes an enormous bright flash under the surface, with a very spectacular effect. The school, which usually travels at from 2 to 7 knots (3.7 to 1 3 km/lir), leaves a rather milky trail behind it, due to the emission of sexual products. H. Maturity and Fecundity The ages at which bluefin tuna spawn, and the variation in the fecun- dity of females with their size and age, are two elements of the biology of the species which are especially important for the management of its fisheries. 1. Age and Size at First Maturity: Research on bluefin tuna in the eastern Atlantic and Mediterranean in- dicated that they first spawned at ages 2-4 (lengths of about 75-125 cm, weights of about 12-40 kg). Sella (1929a, 1929b) reported that the spe- cies usually spawned first at age-3, or at a weight of about 15 kg. He added that a few age-2 fish which had at- tained weights of 12 kg through rapid growth might also spawn. Frade and Manafas (1933) found that young bluefin 1 m long taken in the "arrival" (May-June) fishery off southern Portugal developed difterently according to their sex: the males were in active or completed spermatogen- esis, whereas the females were In a very retarded ovogenesis. Arena ( 1 964) and Sara (1973) have ob,served that bluefin tuna school by size, even dur- ing the spawning season, and that fer- tilization is accomplished by paired emissions within these schools. There- fore the reported difference in the times of maturity of 3-year-old males and females implies that their spawning ef- forts might be ineffective. Rodriguez- Roda (1929a) concluded from his fe- cundity studies that females attained their first maturity at a length of about 97.5 cm, and males at about 105 cm, corresponding to ages of 3 Nears. Frade and Vilela (1962) concluded that first maturity usually occurred at age 3, but occasionally at ages 2 or 4. Scaccini et al. 1975) reported the smallest mature bluefin tuna observed by any of them was an individual weigh- ing 27 kg (4 years old) examined by Sara at the Solanto trap in Sicilv in mid-June 1959. 92 Less information is available on the age of first spawning of western Atlantic bluefin, but these fish have apparently been less precocious than their eastern Atlantic and Mediterra- nean counterparts. Westman and Neville (1942) ex- amined many small and medium sized bluefin tuna landed by the coastal sport fishery at Freeport, Long Island, New York. They reported that nearly all of the "school tuna" (fish averaging less than 30 kg) were immature. They found some evidence of approaching matu- rity in a few tuna of the 3 -year-old age group, and signs of maturity in a larger percentage of the 4-year-olds. They con- sidered all of the fish of both sexes which were 5 years old (about 1 30 cm long and weighing about 40 kg) or older to be adults, but found no indica- tions of eggs or sperm in their gonads. The present authors have con- ducted macroscopic examinations of the gonads of numerous small bluefin tuna caught in traps or by sport fishing in southern New England coastal wa- ters and by purse seines in coastal wa- ters between southern New Jersey and New England in the summers of 1 950- 1975. We concur with Westman and Neville's conclusions. We found signs of maturity in less than 1% of the 3- year old fish but in a considerably larger percentage of the tuna of age group 4. Nearly all age 3 fish, like all the 0, 1 , and 2 year-old individuals examined, had very small gonads which were al- most uniformly slender throughout their length, like very flat shoestrings. The gonads of the older fish which were considered to show signs of maturity were distinctly enlarged for at least part of their length. The testes of a few age 4 males taken in Cape Cod Bay, Mas- sachusetts, in early July 1953, contained milt. Those of a 4-year-old specimen, 110 cm long and weighing 25 kg, weighed 27.5 g and contained milt. A photomicrograph (970x) taken by R. F. Vaccaro, an Associate Scientist of Woods Hole Oceanographic Institution, showed that this milt contained fully developed sperm. The gonads of most of the 5-year-old tuna (about 133 cm long and weighing about 45 kg) were generally better developed, and most of these fish had probably spawned. We believe that the mature mem- bers of the "small" bluefin group (less than 120 cm long) spawn in ofTshore waters north of the Gulf Stream, along with fish of the "medium" size group, in May and June. As noted in Section VD3, a few small bluefin and several medium sized ones taken by longline in this area and season were examined. The results of macroscopic examina- tions of their gonads were as follows: Age III. Two fish were examined. Both were classified as immature. Age 4. Three males and three females were examined. One fish of each sex showed definite signs of maturity. Age 5. One male was examined. Some milt was squeezed from its testes. Age 6. Twelve fish (six of each sex) 143.5-157 cm long and probably of this age were examined. Two of the females were classified as spent. The other ten fish appeared to be ap- proaching spawning condition. We conclude tentatively from all of these data that the first spawning of bluefin tuna in the eastern Atlantic oc- curs at age 3, in exceptional cases, and more frequently at age 4. By age 6 all, or nearly all, of the fish are spawners. Probably the first spawning of western Atlantic bluefin tuna occurs most fre- quently at age 5, but more research is required to establish this. 2. Fecundity Rodriguez-Roda (1967a) found that the estimated fecundity of bluefin tuna caught near Cadiz, Spain, in May and June generally increased with size offish, from 5.2 x 10" for an individual 130.5 cm long and weighing 54 kg to 32.2 x 10'' for one 230 cm long and weighing 235 kg The minima and maxima, however, were 5.0 x 1 0'' for a fish 160 cm long and weighing 96 kg, and 45.9 X 1 O*" for one 214cm long and weighing 191 kg. He calculated the following relationships between F, tlie fecundity or number of maturing eggs, and (1) L, the length of the fish in cm; (2) and (3) P,, the weight of the fish in kg; and (4) P^, the weight of the ovaries in kg: (1) F = 2.29245 L^ (2) F = 53,451 P^ii""*' (3) F = -1,220,71 7+ 138,068 P, (4) F = 553 P, ' "'"" In the size range studied, the esti- mated fecundity was thus roughly pro- portional to the weight of tlie fish, or the third power of its lengtli. Baglin (1976), studying six blue- fin tuna taken near Bimini and Cat Cay in the northwestern Bahamas in May and June and ranging from 222.5 cm in length and 1 88.4 kg in weight to 260.6 cm in length and 271.5 kg in weight, found that their estimated fecundity in- creased from 16.7 x 10" for the smallest to 3 1 .4 x lO" for the largest. The maxi- mum, however, was 33.0 x lO' for an individual 240.8 cm long and weigh- ing 247.4 kg. Where F is the estimated fecundity, L is the length of the fish in cm, W is its live weight in kg, and W^ is the dry weight in g of all of the eggs from both of its ovaries, Baglin calcu- lated that: F = 65.421 4 L"'"" F = 6,245,010 + 95,132.3 W F- 3,051,104+ 18,916.4 W,. He concluded that fecundity increased with size of fish in the length range from 222.5 cm and to 260.6 cm, and that, therefore, larger fish contribute more to the reproductive potential. No significant difference in the slopes or adjusted means was found by Baglin in a covariance analysis between the fecundity-length and the fecundity- weight relationships found in his study of western Atlantic bluefin tuna and those found by Rodriguez-Roda (1967a) for eastern Atlantic individu- als. Baglin noted, however, that the samples were small, and that more ex- tensive studies might reveal significant differences. The studies of Rodriguez-Roda (1967a) and Baglin (1976) indicate clearly that the fecundity of Atlantic bluefin tuna increases with size offish, in both the eastern and western parts of the Ocean. Baglin's study shows that this relationship extends to the very large size of 260 cm. Pending actual determinations of the fecundity of larger fish, there is no reason to suppose that the fecundity does not also increase with size of fish for the relatively few larger individuals which are captured. Baglin and Rivas (1977) estimated the fecundity of 1 7-year-old Atlantic blue- fin as 44.6 x 10" eggs. 93 3. Maximum Size at which Bluefin Tuna Spawn The only authors who have dis- cussed the largest size or age at which bluefin tuna spawn, to our knowledge, are Scaccini et al. (1975). They state that the species spawns "up to the maxi- mum weight and the maximum age" and cite Sari's observation of bluefin tuna estimated to be about 1 8 years old and weighing up to 600 kg, caught in the trap at Favignana (Aegades Islands, just west of the western end of Sicily) in June 1974, as the largest individuals in spawning condition which had been examined by any of them. Since the nimiber of recorded catches of bluefin tuna weighing over 600 kg is negli- gible, it is most unlikely that any sig- nificant number of these fish attain a size at which they cease to spawn. I. Discussion and Conclusions A discussion and a presentation of our tentative conclusions seem appro- priate, since much of the information in this section is inconclusive and, in many instances, even contradictory. 1. Mediterranean and Black Seas The Mediterranean and Black Seas have traditionally been regarded as prime spawning grounds for the blue- fin tuna. Despite a preponderance of scientific opinion against this concept which developed in the last quarter of the nineteenth century and the first quar- ter of the twentieth, it still appears to be valid. The information now available, however, indicates that the principal reproduction of the eastern Atlantic and Mediterranean bluefin takes place in the south central Mediterranean, instead of occurring in the Black Sea as Aristotle and his followers supposed. Small juvenile bluefin were first reported from the Mediterranean by d'Amico (1816). The eggs and larvae were first described briefly by Sanzo in his 1910b publication, and more com- pletely in his 1929 and 1932 works. Some authors have stated that the char- acters presented by Sanzo do not dif- ferentiate the eggs and very small lar- vae of bluefin from those of other tuna- like fishes, especially Auxis. Ehrenbaum's tentative identifications of larval T. thynnns thynnus have been questioned by Sella (1929a) and Richards ( 1 976). Despite these uncer- tainties over identifications, there is no doubt that young stages of bluefin have been collected in great numbers in the south central Meditenanean. It is un- fortunate, as noted by Heldt ( 1 930) and Dieuzeide (1951), that Sella did not describe the "thousands" of early stages of bluefin which he (1929a) reported that he had collected. His 1924 pub- lication, however, convinced Richards (1976) of the accuracy of his identi- fications. Scaccini (1961, 1968) and Scaccini et al. (1975), to whom Sella's material was available, also agreed with these identifications. The most numer- ous collections have been made in the Strait of Messina and in waters north and west of Sicily (Sanzo 1932, Sella 1924, 1929a; Sparta 1933, Scaccini 1968, Piccinetti and Piccinetti Manfrin 1970, Scaccini et al. 1975). The de- tailed study of Duclerc et al. (1973) showed that larval bluefin also occurred around the Balearic Islands. Dieuzeide (1951) described three larval bluefin collected off Algeria, and Piccinetti ( 1 973) and Piccinetti et al. ( 1 976b) pre- sented preliminary reports on occur- rences of early stages in the Adriatic Sea. Several authors (Vodyanitskii 1936, Vodyanitskii and Kazanova 1954, Oven 1959) have reported on eggs and larvae of bluefin collected in the Black Sea, but have not produced detailed descriptions. The difficulties of identifying eggs and small larvae of bluefin, even after hatching the eggs and rearing the larvae, have been fully explained by Duclerc etal. (1973) and Scaccini etal. (1975). The regular occurrences of great numbers of maturing bluefin along the coasts of westem Sardinia, Tunisia and Tripolitania suggest that spawning oc- curs in those areas as well Although the traps on the eastem (Ionian Sea) coast of Sicily were known as "return" traps, some maturing individuals were caught in them, along with the more numerous spent fish (Sella 1929a, Scordia 1938). Scordia's extensive stud- ies (summarized in her 1938 and 1942 publications) indicated that large num- bers of bluefin tuna passed through the Strait of Messina from the lyrrhenian Sea to the Ionian Sea, spawned there, and then returned through the Strait to the Tyrrhenian. It appears that the spawning of bluefin tuna in the Mediter- ranean is most intense in its south-cen- tral part, but also extends into the west- em Mediterranean and the Adriatic Sea. As Scaccini et al. (1975) concluded, additional research will probably show that bluefin spawn in other parts of tlie Mediterranean. Some spawning must also occur in the Black Sea. The ab- sence of a commercial fishery for the species there, however, and the rela- tively small catches taken in its ap- proaches, the Bosphorus and the Sea of Marmara, suggest that the reproduc- tion there is not comparable, quantita- tively, to that in the central Mediterra- nean. 2. Eastern Atlantic The situation in the eastem Atlan- tic is dramatically different from that in the Mediterranean and Black Seas. De- spite extensive research efforts over many decades, no identifiable early stages of bluefin have been collected in the Ibero-Moroccan Bay, which has been regarded as the prime spawning ground in the region, or in the more recently suggested areas of reproduc- tion off Morocco and in the Bay of Biscay. The only identified early stages from the eastem Atlantic were collected near the Equator between longitudes 0° and 15=30'W (Richards 1969, Richards and Simmons 1971). This was a totally unexpected area on the basis of existing knowledge of the distribu- tion of maturing bluefin. It certainly merits further investigation to deter- mine the seasonal and areal extent of the occurrence, and also to make cer- tain that the larvae are actually those of T. thynnus ihynnus rather than those of T maccovll, which also occurs in the South Atlantic. Regular seasonal occurrences of very numerous maturing and spent blue- fin in the Ibero-Moroccan Bay have been abundantly documented. Small numbers of maturing or ripe bluefin in the Bay of Biscay have been reported occasionally (Creac'h 1952, Le Gall 1952) and detailed studies of the go- nads of a few individuals have just become avai lable (Cort et al. 1 976, Cort 1977). The occurrence of small num- bers of mature bluefin at the Azores has been reported (Ferreira 1932), but no details on their gonad condition were presented. 94 3. Western North Atlantic The western North Atlantic is the only part of that ocean from which extensive collections of early stages of T. thyiinus thynnus have been obtained. The most important occurrence docu- mented up to now has been in the Gulf of Mexico. Bluefm larvae and small juveniles have been collected over much of the deep (more than 300 m) part of the Gulf, especially in the area north of latitude 23 °N (Judrez 1972 1 974b; Richards 1976, 1977;Montolio and Ju^ez 1977, T.C. Potthoff, per- sonal communication). Important col- lections have also been made in the Straits of Florida, particularly at the Dry Tortugus and off Miami (Potthoff and Richards 1 970, Richards 1 976, T.C. Potthoff, personal communications). Scattered individuals have been col- lected off the east coast of the United States up to latitude 38°45'N. These collections show that the Gulf of Mexico is the most important bluefrn tuna spawning ground yet discovered in the Atlantic, and that spawning also occurs in the Straits of Florida and prob- ably for an unknown distance farther north. The condition of gonads of cap- tured fish suggests that spawning may also take place in the northwestern Car- ibbean, the Windward Passage, the Old Bahama and Santaren Channels, and a large area east and north of the Baha- mas. Virtually all of the mature bluefin taken in these southerly parts of the western North Atlantic and adjacent waters for which we have size data were large (over 185 cm long) indi- viduals, weighing over 125 kg. Exami- nations of gonads suggest that at least some smaller bluefin spawn near the northern edge of the Gulf Stream, or the Gulf Stream front, north of latitude 37°N and east of longitude 70° W. 4. Overall Situation On the most positive evidence available, the occurrence of larvae and very small juveniles, one can only con- clude that the most important spawn- ing areas of T. thymus thynnus are in the south central Mediterranean Sea and the Gulf of Mexico. This in itself is an interesting parallel, as the Gulf of Mexico and the Caribbean Sea are of- ten referred to as the "American Medi- terranean." The major spawning occurs ear- lier in the western Atlantic (probably about May 15-June 15) than in the Mediterranean (about June 1 5-July 15). The smaller bluefin spawn later than the larger ones in the Mediterra- nean, and probably do in the western Atlantic also. Studies of spawning in both areas have centered on the larger fish. Some information is available on the spawning of the smaller individuals in the Mediterranean, where they mix to a considerable extent with the larger ones during the spawning season. The two groups evidently reproduce sepa- rately in the western Atlantic, and much less is known about the spawning of the smaller fish (all of the "medium" size group, and the mature members of the "small" group) there. The only occurrence of early stages of bluefin in the eastern North Atlantic, or the South Atlantic of which we have knowledge is the collection of a very few larvae near the Equator and longi- tude 7°W. Much more information will be needed to determine the significance of this unexpected finding. Other collections of larvae, al- though less important than those from the Gulf of Mexico and the central Mediterranean, indicate that bluefin spawn around the Balearic Islands, in the Black Sea, in the Straits of Florida and, probably, north of the Bahamas. Studies of the maturity of gonads suggest spawning over more extensive areas. Ihese would include major spawning along the coasts of Tunisia and Tripolitania, and in the Ibero-Mo- roccan Bay (by the fish which do not enter the Mediterranean to spawn). Spawning in the Bay of Biscay is also indicated but this is probably less im- portant. Gonad studies in the western Atlantic suggest, extensive additional spawning areas off Cuba, the Baha- mas, and the southeastern United States along the Gulf Stream front north and east of Cape Hatteras. Judrez (1974b) and Montolio and Juarez ( 1 977) provided quantitative es- timates of the numbers of bluefin lar- vae in extensive areas of the Gulf of Mexico. Richards ( 1 976) estimated the number offish in the spawning stock from the estimated number of larvae (Juarez 1974b). His figure was in rea- sonable agreement with those calcu- lated by other methods. No quantitative estimates of the abundance of bluefin tuna larvae in the Mediterranean have been published. C. Piccinetti (personal communication), however, informed the senior author that the relative abundance of larvae in the Gulf of Mexico, as indicated by Montolio and Judrez ( 1 977), was much less than that which he and his col- leagues had found in parts of the Medi- terranean. 95 VI. MIGRATIONS AND STOCK IDENTIFICATION A. INTRODUCTION Identification of Atlantic blue- fin tuna stocks is regarded as one of the most important prerequisites to the efficient management of its fish- eries. The studies of stock identifica- tion and migrations are so closely related as to be almost inseparable. Some methods, such as the visual or electronic tracking of fish, and the analysis of the seasons and localities of catches, relate primarily to migra- tions. Others, such as biometric and biochemical studies, are more directly concerned with stock identification. Tagging is one of the most positive methods of studying both problems. In difficult cases, however, it is nec- essary to use all available means to achieve either objective. This is es- pecially true of the bluefin tuna, whose long, rapid and variable mi- grations make their migratory pat- terns and populations especially dif- ficult to identify. Until 1954, when sustained tag- ging of the Atlantic bluefin was initi- ated, scientists were limited to de- ductive, or indirect, methods of study- ing its migrations. Likewise, until the development of biochemical meth- ods at about the same time, they de- pended mainly on biometric studies to identify its populations. Even with the aid of these and other advanced techniques, the migratory patterns of bluefin tuna are not completely un- derstood and the stocks have not been positively identified. In this section we will consider the migrations and stocks (without a priori implication that they are sepa- rate) in the Mediterranean-eastern Atlantic area and in the western At- lantic. Finally, we will consider the implications of trans-equatorial and transatlantic migrations and present our conclusions in regard to the iden- tity of stocks. B. METHODS AND MATERIALS, AND DEFINITIONS \. Methods and Materials a. Deductive Methods Deductions in regard to migra- tions and the identity of stocks have been based on observations or infor- mation of the following types: ( 1 ) The times of appearances and dis- appearances of the fish in fishing areas and the observed move- ments of the fisheries for it. (2) The size or age composition of the catches. (3) The sex ratio of the catches. (4) The observed behavior of the fish in specific areas and seasons. (5) The apparent preferences of the fish for environmental conditions, and their seasonal changes. (6) Deductions as to the localities where specially constructed hooks and lures found in bluefin tuna caught in traps had origi- nated. b. Methods Used Mainly for Identifying Stocks Just as many of the above meth- ods have been used to identify stocks as well as to study migrations, the following ones, which have been used primarily for stock identification, have also been applied to the study of migrations: (1) Anatomical comparisons. (2) Biometric comparisons of mor- phological characters. (3) Comparisons of biochemical properties. (4) Comparisons of areas and sea- sons of spawning. (5) Comparisons of growth rates. c. Methods Used Mainly for Studying Migrations Most of the methods listed be- low were developed to study migra- tions, but tagging is equally useful for identifying stocks: ( 1 ) Visual tracking from aircraft. [Vi- sual observations from vessels or shore come under item a(4).] (2) Tagging. (3) Tracking with sonar. (4) Tracking with sonic tags. d. Development of Bluefin Tuna Tagging Tagging is the only positive method of determining that an indi- vidual fish has gone from one place to another. The desirability of tag- ging Atlantic bluefin tuna was rec- ognized long ago (Sella 1912b), but the technical difficulties were such that it was not accomplished on a continuing basis until 1954 (Mather 1960, 1963). Sella (1927) tagged 20 bluefin tuna weighing from 4 to 20 kg off Gailipoli (southern Italy) in 1912 with bands around the caudal peduncle, but no returns resulted. His (1927, 1929a) deductions of migra- tions from hooks and lures found in tuna caught in traps, however, aroused new interest in the problem. Methods of marking were discussed extensively at the "Conference of ex- perts — '■ (Anonymous 1932b), and the tagging of bluefin tuna was strongly recommended. Large and small bluefin tuna were tagged off Portugal in the years 1931-1935 and 1960 (Frade and Dentinho 1935, Heldt 1938, Vilela 1960). Some large individuals were also tagged off Tu- nisia and Morocco (Heldt 1943). In addition Heldt (1932) distributed marked hooks to tuna fishermen at Groix, France, on the Bay of Biscay in 1927, 1928 and 1929. Many of the hooks and lures retrieved by Sella (1929a) from tuna which had been caught in Mediterranean tuna traps 96 were of the type ordinarily used in that bay. English sport fishermen, co- operating with F. S. Russell (1934a), left marked hooks in a number of giant bluefm tuna in the North Sea in 1933. Since no returns resulted from all of these efforts, interest in the matter subsided again. The first successful tagging of Atlantic bluefm tuna was achieved by Westman and Neville ( 1 942), who marked 23 small bluefm tuna with celluloid disc tags off New York Har- bor during the 1941 fishing season. They obtained two returns in the same locality and season, after periods of less than 75 days at large. Interest in such matters of course ceased with the outbreak of World War 11. Cooperating sport fishermen used numbered hooks furnished by Schuck and Mather to mark giant bluefm tuna off the Bahamas in 1950 and 1951, and off Rhode Island in 1952 (Mather 1963). Rivas (1954) also used the numbered hook method to mark giant tuna off the Bahamas in 1952 and applied strap tags to the opercles of some of these fish, like- wise utilizing the cooperation of rod and reel anglers. The only result from all these efforts was the return, about fifteen years after its recovery, of a numbered hook from a fish recap- tured from Wedgeport, Nova Scotia. The release data for this hook, which had been sold by a tackle dealer in eastern Long Island, New York, had not been reported, and our efforts to retrieve this information were unsuc- cessftil. The fish probably had been released, or broken free, off eastern Long Island, but of course this is uncertain. Interest in tuna tagging was re- vived when the California Depart- ment of Fish and Game developed the dorsal loop tag and used it suc- cessftilly on small Pacific tunas (Wil- son 1953). Mather, with the coopera- tion of a few interested sport fisher- men, proved the feasibility of mark- ing small Atlantic bluefin with the dorsal loop tag in Massachusetts wa- ters during the 1954 fishing season. He concurrently developed the dart tag, with which even very large fish could be marked rapidly once they were brought alongside the boat (Mather 1960, 1963). The successful Cooperative Game Fish Tagging Pro- gram of the Woods Hole Oceano- graphic Institution, which has been conducted jointly with the National Marine Fisheries Service since 1974, developed from these beginnings. Subsequently agencies in other At- lantic and Mediterranean nations have marked bluefin tuna and other large pelagic species with interesting and important results. e. Summary of Bluefin Tuna Tagging Programs A summary of the more impor- tant programs, with the major areas and dates of operation and major lit- erature references for each, follows: Canada — Fisheries Research Board of Canada (now Environment Canada). St. Andrews Biological Laboratory — Northwestern Atlantic- 1963- present, Beckett (1970), Burnett etal. (1977). France — Institut Scientifique et Technique des Peches Maritimes- Bay of Biscay and Portugal- 1 968- present, Aloncle (1973). Italy — Centro Sperimentale della Pesca-Tyrrhenian Sea-1962- present. Arena and Sar^ (1967), Arena and Li Greci (1970). Morocco — Institut des Peches Maritimes du Maroc-Atlantic coast of Morocco- 1972- 1 973, Lamboeuf(1975). Norway — Fiskeridirektoratets Havforskningsinstitutt-West coast of Norway-1957-1962, Hamre(1965). Spain — Instituto de Investigaciones Pesqueras — Cadiz area, Spain- 1960-1967, Rodriguez-Roda (1963, 1964c, 1969a). United States of America — Woods Hole Oceanographic Institution and National Marine Fisheries Service — Northwestern Atlan- tic and adjacent waters- 1954- present and 1 974-present, respec- tively, Mather (i960, 1962, 1969, 1974a), Mather et al. (1967, !974b), Mason etal. (1977). The results of tuna tagging have been summarized and briefiy de- scribed by the FAO Panel of Experts for the Facilitation of Tuna Research (1972), Mather and Mason (1973, 1976), Mason (1975), and Mason et al. (1977). 2. Definitions Direct migration-This term is used to designate the movement of a fish which presumably could not have been repeating an annual migratory pattern when recap- tured. It does not imply that the fish has travelled on a straight course from one point to another. Trap fisheries-Special terms of the trap fisheries have been defined in Section 1IB4. 3. Hypothetical Migration Model Subject to some variations be- tween successive ages and overlaps between age groups, and some re- gional differences, we believe that the basic migratory behavior of the bluefin, through its life in the Atlan- tic, may be summarized by age groups as follows: Very Small Fish (Less Than 2.5 kg, and Age 0) — Development from egg to active predator is ex- tremely rapid. Hatching occurs within two days. The larvae be- come active swimmers in 1 5 days, and a length of 30 cm may be attained within three months. The first important movement of the newborn fish is one of concen- tration. They migrate from ex- tensive spawning areas to lim- ited nursery (feeding) grounds. Growth during their first winter is much slower than it was dur- ing the warm season. Small Fish (2.5-32.0 kg, Ages 1-4, Immature) — These fish make annually repeated two-phase mi- grations between limited warm season coastal nursery areas and little known, but presumably more extensive, cold season win- tering areas. The warm season occurrence is typically in the sur- face layers, with heavy feeding and rapid linear growth, but no decrease in the length-weight ra- tio. The winter sojourn, on the other hand, occurs at deeper lev- els, and growth virtually ceases. 97 Consequently, feeding and gen- eral activity are presumed to be greatly reduced. The migratory behavior of mature individuals in this group often resembles that of the medium fish. Medium Fish (32-122 kg. Ages 5-8, Mature) — These fish make an- nually repeated three-phase mi- grations between warm season feeding areas, deep water win- tering areas, and late spring spawning areas. The warm sea- son distribution is more exten- sive than that of the small fish, and not as limited to coastal wa- ters. This occurrence is again fre- quently in the surface layers, but not as predominantly so as that of the small fish. Feeding is heavy. Seasonal linear growth has not been determined, but is pre- sumably rapid. The length-weight ratio decreases considerably dur- ing this period. The cold season distribution is much wider than that of the small fish, extending far into oceanic waters. These fish then remain in the subsurface lay- ers and it is assumed, but has not been demonstrated, that their feeding, activity and growth are reduced. In late spring and early summer, many of these fish con- centrate in spawning areas which are little known. After spawning they return to the warm water season feeding areas. Large or Giant Fish (Over 122 kg. Age 9 or Older, Mature) — The fish make three-phase migrations basically similar to those of the medium size group, but their dis- tribution and migrations are much more extensive. Their spawning and migratory periods, as well as the areas concerned, while dif- fering somewhat from those of the medium group, overlap or co- incide with them to a consider- able degree. The seasonal linear growth of these fish is not known, but their length-weight ratio de- creases greatly during the feed- ing season, and increases corre- spondingly during spawning and their post-spawning migrations. As with the smaller groups, their feeding, general activity and growth are presumed to be re- duced during their wintering pe- riod in deep waters. Movements within some of these seasonal habitats have been observed, as well as the migrations between them. We shall discuss the migratory patterns of the different size groups of bluefin tuna in terms of this sim- plified model. C. STUDIES OF MIGRATIONS AND STOCK IDENTITY 1. Mediterranean and Eastern Atlantic a. Introduction The earliest speculations and hy- potheses about the migrations of blue- fin tuna, and in fact nearh all of those before 1920. were concerned with the Mediterranean and eastern North Atlantic areas. In fact, the study of these migrations undoubtedly de- veloped from observations made dur- ing fishing operations in the Medi- terranean and its approaches during the pre-Christian era. Traps similar to those still in use apparently ex- isted then, but movable nets, whose successful operation required the assistance of watchers (thynnoscopi) situated at coastal vantage points, were probably more numerous (Parona 1919, Thomazi 1947). Both methods depended on the movements of schools of fish along the coasts, which occurred mainly during the runs of maturing (now known as "ar- rival") tuna in May and June, and post-spawning (now known as "re- turn") fish in July and August. There- fore, before recapitulating the devel- opment of hypotheses on the migra- tions and populations of tuna in this region, we will summarize some of their local movements which have been indicated over the ages by the operations of the traps which have harvested these periodic passages. Basic data which will be considered include the location of the traps, the direction from which fish must ap- proach them, and the periods when they catch tuna. The relative impor- tance of their catches will be dis- cussed in special cases. Sara ( 1 964, his Figure 5) showed the approximate locations of the significant traps still in use in the early 1960s, indicating, for each, the fishing period and the direction from which the fish entered the trap. The data for those in the Ibero-Moroccan Bay and the extreme western Medi- terranean (Figure 45) strongly sug- gested eastward movements from the Atlantic into the Mediterranean ("ar- rival") in May and June, and move- ments in the opposite direction ("re- turn") in July and August. The arrays of "arrival" traps on the sides of the Bay terminated at Tarifa (north coast) and Cape Spartel (east coast), show- ing that the "arrival" run extended at least up to the very threshold of the Strait of Gibraltar. Data for the "return" traps sug- gested that some west-bound tuna fol- lowed botli coasts of the Mediterra- nean as the\ approached the Strait. In the Ibero-Moroccan Bay, however, they occurred in much greater strength along its northern (Iberian) coast, in contrast to the complete ab- sence of a "return" trap fishery along its eastern (Moroccan) coast. The other important group of traps shown by Sara (1964) was in the central Mediterranean. Most of these fished the "arrival" run only. The indications from the dates and directions in which the fish entered the traps were as follows. Off Tunisia, the "arrival" tuna apparently travelled eastward from Bizerte to the Cape Bon peninsula, which they rounded, then moved southward to Ras Kapudia. "Arrival" bluefin also evidently moved east- ward along the Libyan coast, from Zuara near the Tunisian border to Cape Misurata (longitude 15°E). The situation around Sardinia, Sicily, and Italy differs because of the disjointed configuration of the coasts. The "arrival" fish tended to move in southerly directions along the western shores of Sardinia, and also along the Calabrian coast in the Gulf of Sant' Eufemia. With local exceptions at its two extremities, "ar- rival" fish travelled westward along the north coast of Sicily, following the western sides of bays and the eastern sides of promontories along the way. This westward movement is 98 directly opposite to the usual "ar- rival" pattern. The "return" situation is much simpler. Tuna apparently swam southward along the southern half of the east coast of Sicily, and west- ward along the western half of its southern coast. The above discussion is obvi- ously simplified, and is only intended to show trends in the movements of the tunas. There have been many varying opinions on how the tuna approach the coasts and enter the traps (Scaccini and Pacagnella 1965), We do not mean to imply that the entire mass of fish in a given area follows the route described, but only that many fish in the respective areas tend to travel in the stated directions. Most of the early theories on the migrations were based mainly on in- formation of the type which has been summarized above. It should be noted, however, that the European fisheries have a very ancient history, whereas most of those in Africa origi- nated in modern times. The theories on migrations of bluefin, developed largely from stud- ies of the trap catches, and later from consideration of environmental fac- tors as well, varied from Aristotle's (circa 325 B.C.) view that the bluefin was an Atlantic species which occu- pied the Mediterranean only on its way to and from its supposed spawn- ing grounds in the Black Sea, to the concepts of Pavesi (1889), Sanzo (1910a), Roule (1914a), F. de Buen (1925), and Scordia (1938) that the Mediterranean and Atlantic bluefin constituted entirely separate popula- tions. Pavesi (1889) thought that the Mediterranean bluefin was an abys- sal animal, rising from the depths of that sea and moving to nearby sur- face waters only to spawn. Bounhiol (1911) shared in the concept of a strictly Mediterranean bluefin stock but believed that its apparent migra- tions were actually the result of a tendency to swim against the wind- driven currents. Roule (1914a) felt that the bluefin was a pelagic crea- ture which migrated only within the basin of the Mediterranean in which it lived. He believed that its spawn- ing migrations were controlled by the temperature and salinity of the wa- ter. Scordia (1938) and others (Ninni 1922, Genovese 1957) thought that there were two or more stocks of bluefin within the Mediterranean, each native to its particular basin. Some authors even questioned whether bluefin tuna actually oc- curred in the Atlantic (except in the Ibero Moroccan Bay) in significant numbers, or whether the bluefin re- ported from the Atlantic were of the same species as those in the Mediter- ranean. These theories were predom inant in the late 1800s and early 1900s, until Sella (1926a, 1926b, 1927, 1929a, 1929b, 1930, 1932a, 1932b) hypothesized several migrations from various Atlantic areas and from the Sea of Marmara into the Mediterra- nean, as well as between the differ- ent basins of that sea. He deduced these migrations by determining the localities where hooks and lures found in bluefin, which had broken lines or leaders and subsequently been caught in traps, were in general use. In that period, tuna fishermen used hooks and lures which were hand- made locally. Distinctive designs and methods of attachment to lines or leaders made hooks and lures from different localities easily recogniz- able. Thus it might be assumed that a fish found carrying a hook or lure typical of a given area had come from that area. In some cases, this prob- ability was greatly increased by many findings, indicating similar migra- tions. This somewhat uncertain method was followed by the more positive ones of sonar tracking (LozanoCabo 1959a, 1959b) and tag- ging (Rodri'guez-Roda 1964c, 1969a; Hamre 1965, Arena and Li Greci 1970, Aloncle 1972, Lambouef 1975). Studies of the effects of environmental conditions on the oc- currences and behavior of tunas, par- ticularly during the spawning sea- son, and analyses of the time, loca- tion, quantity and size composition of catches, continued to provide de- ductive indications in regard to mi- grations concurrently (J. Le Gall l929,J.Y.Le Gall 1974. Hamre 1958, 1962, 1965; Lozano Cabo 1958, Rodriguez-Roda 1963, 1964a, 1969b, 1970a, 1970b; Aloncle 1964, Arena 1964, Sara 1964. 1973;Tiews 1964). Numerous biometric compari- sons of morphometric and meristic characters of bluefin tuna from vari- ous areas were carried out to identify populations or races (Frade 1931, Arico and Genovese 1953, Nedelec 1954, Genovese 1957, 1958). More recently, some genetic and biochemi- cal research has been undertaken for the same purpose (Keyvanfar 1962, Lee 1965, 1968). Despite all of these efforts, un- certainties about the migration pat- terns and the identity of the stocks in the eastern Atlantic and the Mediter- ranean still exist. b. Migrations Between the Mediterranean and the Eastern Atlantic Unit stock or two stocks? — The question of whether the bluefin tuna constitute a single stock which mi- grates from the Atlantic into the Mediterranean to spawn and then out again, or comprise two stocks — one Atlantic and one Mediterranean — has been debated for decades. This discussion led to such terms as the "migratory" or "sedentary" tuna, (re- ferring to the supposed habits of the fish) and the "unitists" or "dualists" (referring to the theories on the num- ber of stocks). Ancient trap fisheries in the Mediterranean Sea and its approaches were based on the spawning runs of the larger bluefin: an eastward "ar- rival" run. mostly of fat fish with ripening gonads, in May and June and a westward "return" run, mostly of lean spent fish, in July and Au- gust. These runs were the basis of the Aristotelian theory, that the bluefin was essentially an Atlantic species, but passed through the Mediterranean in May and June to spawn in the Black Sea, and returned to the Atlan- tic in July and August. This theory has been accepted, wholly or in part, by nearly all of the trap fishermen to this day, and also by many scientists. Objectors (Pavesi 1887, 1889; de Bragan^a 1899, Sanzo 1910a, Roule 1914a, 1914b, 1917, 1924; F. de Buen 1925, 1931; O. de Buen 1924, Scordia 1938) noted that blue- fin spawned in the Mediterranean and that individuals of all sizes were present in it throughout the year. They also maintained that the lag which 99 would have occurred between catches in "arrival" traps located in westerly areas and those in easterly areas, if all were fishing a single group offish moving eastward, did not exist. Roule and F. de Buen used the relatively poor bluefin catches in the Sea of Alboran (just east of Gibraltar) as another argument against a large scale spawning migration of bluefin from the Atlantic into and out of the Medi- terranean. Both of these scientists cited the sensitivity of the bluefin tuna to environmental conditions as evidence that they could not pass through the relatively cold (R. de Buen 1927) waters of the Strait of Gibraltar, especially during the spawning season. Roule (1917) pro- posed the "halothermic" theory, maintaining that the bluefin were "stenotherms" and "stenohalines", and sought the warmest and most saline waters during the spawning season. F. de Buen (1925) opted for the "thermic" theory, considering water temperature to be the domi- nant factor controlling the tunas' movements. Both agreed, as did most students of the subject in the Mediter- ranean and adjacent Atlantic waters, that the sensitivity of the bluefin tuna to the environment increased during the period of maturation. Scordia (1938) also believed in the "sedentary" tuna theory and felt that few, if any, bluefin tuna entered the Mediterranean fi-om the Atlantic. Like Roule (1914a, 1914b, 1917) and F. de Buen (1925), she stressed the increased sensitivity of the maturing tuna to its environment. Sella ( 1 929a) showed, however, that bluefin tuna could withstand ex- treme changes in temperature and salinity during their feeding period. He maintained that even during the spawning period they sought specific conditions which varied with size of fish, rather than maxima as proposed by Roule (1917). He further argued that the "thermic" or "halothermic" barrier of F. de Buen (1925) and Roule (1917) did not constitute proof that bluefin could not pass through the Strait of Gibraltar, but would be an explanation of such a situation, if its existence were substantiated. Sella noted, as Sara (1964, 1973) and Scaccini et al. (1975) did later. that there were relativeh few large bluefin in the Mediterranean except during the spawning migrations and that the arrival fishery actually did begin later in the season in propor- tion to how far east the traps were located. After this long period of indirect studies, Sella's (1927, 1929a) find- ing of 25 Atlantic hooks and lures in tuna caught in Mediterranean traps revived the "migratory" or "unit stock" theory. The items recovered included 13 from Tarifa, seven cer- tainly and two probably from the Bay of Biscay, two from the Azores, and one from south of Ireland. These star- tling revelations did not pass unchallenged in this period when the "sedentary tuna" theory was deeply rooted in the minds of nearl) all the scientists concerned with the prob- lem. F. de Buen (1931) and Scordia ( 1 934) questioned Sella's deductions, on the grounds that fishermen from the localities where the hooks and lures were made might actually have used them while travelling in other areas. Although they admitted that a few such migrations might occur, they maintained that they probably were not related to the spawning cycle, and were relatively insignifi- cant numerically. Sella (1929a) had already pointed out, in support of his hypothesis, that the findings of the hooks varied according to the inten- sity of fishing in the area of their origin. When the fishery off Tarifa was important, numerous hooks of the very distinctive type used there were recovered in the Mediterranean. After the fishery off Tarifa had de- clined drastically, and the one in the Bay of Biscay had increased greatly, hooks of the former area were no longer found, but hooks of the Biscay type appeared in numbers. Sella also noted that, considering the enormous odds against recovering a lost hook in a tuna, the numbers of findings indicating these migrations were too large to be attributable to casual fish- ing by transients. Detailed information on the movements of bluefin tuna in the Strait of Gibraltar obtained by echo sounder was reported by Lozano Cabo ( 1 959a, 1 959b). In 2^000 miles of cruisina in the Strait, and as far as Barbate and Cadiz, Spain, and Larache, Morocco, in June and July 1 957, he obtained data which he con- sidered sufficient to provide interest- ing conclusions on the biology of the species. Lozano Cabo (1958) had sus- pected that bluefin, at least during their spawning migration, might be negatively rheotactic (tending to swim with the currents). The detec- tion of schools confirmed this. The greatest number of "arrival" schools were found in the middle of the Strait, where the current always favored their eastw ard passage. Many schools were also found on the northern side, where the current was usually favor- able, but they were very rare on the Moroccan side, where the currents tended to vary with the tides. The depth at which the schools traveled was difficult to determine, but ap- peared to be somewhat less than 60 m. The migration route of arrival at the Strait when departing from the Atlantic followed the Spanish coast. Some schools uere located between Cape Spartel and Cape Malabata (on the southwestern side of the Strait), but none were detected south of Cape Spartel during numerous cruises be- tween Tangier and Larache. Perhaps the tuna which came from the Afri- can coast partially crossed the Strait from off Cape Spartel toward Tarifa, and joined those coming from the Spanish coast, or followed routes similar to theirs through the Strait. The schools which had tended to con- centrate in the center of the Strait while traversing it appeared to dis- perse as they entered the Mediterra- nean at the Ceuta-Algeciras line. In general, the migrating tunas appeared to prefer temperatures of 18X to 21°C, but deviations from the preferred temperature were more frequent in the Strait than at the traps. Lozano Cabo attributed this to the overcoming of temperature sensitiv- ity by the reproductive urge. Nearly all previous scientists (Roule 1914a, Sella 1929a, Scordia 1938) had main- tained that the sensitivity of the blue- fin to temperature and salinity was greatly increased during the spawn- ing period in contrast to the "erratic" (feeding) period. In an\ case, Lozano 100 c o a. 13 ic o T3 .5 ci. on T3 U 3 TO T3 cs H e .d J- 2 § u u U O a. OS < S V s U O t/3 • — i s 2 ^ Oi w — V OS < 2 E U 3 Qfi Z p — ; O (N oop'^vq~r) .o OS ^- I I , OsOS — ,' ^^ ', QsQSQS ^^^UQ^^^^^^^^^^^^ UJ u ^ O r-- so ^ d ■ri so (^ so rn CN o u-> u-i ^- m o m o o so O so — — f-i »y-l rn o o r^ t^ >0 O O O SO so O O o r) oi o — — — mm O — — o o ZZZZZZZZZZZZZZ o OS r- o o o i^ so r- m (^ f*^ o o rs| m m — f^ so "O TT m m o o o o o o o O r^ O ro rn r*^ r- "o >o r*~) r~i m r- — (-. m ro m O oo oo Tr\ \r\ ra nil ii fin o nn 3 c C c 3 c '(3 i o Q. a. t- o 5 a. o Cl a. OD C/3 u. a. C/5 a. en 00 3 t O a. o o 2 S 03 o w nil (J nil o c c C o t: « ,*s o >-s o n ej 2 o u. Q. c a. oo _1 _l oo o o o — — SOSOO^SOOSOS^OSOS OSOS--OS-7 — __—■-- O O so SO (N rsi so SO OS OS _„ , ^ OS _ so OS 00 SO OS >>>>>>>>>>>>>>>> I I w so so so — >/^ 1/-1 so in u-i 000 so so so n >n 000 so so so >n >n w-i in 000 so so so m >n "n m m m 000 so so m >n >n m o o so so v-i >n >n >n o o so M vn m o — ZZZZZZZZZZZZc/)Z z z z z (N OS OS ra 13 X) X (3 13 X X 13 13 X X Cdc3(3flc3c3c3c3c3c3c3 CB 13 X X C3 CQcacQCQCQCQCQCQtnCQfflCQaQCQfflCQ 2 =3 -e c C3 3 CO a. — fNm'3->nsor^ooos O — r-)r>^Trmsor~ooos X -a c SI 3 o T3 u r. o ex > o o ■o t •^ 2 13 O X c C/5 rt> CO (/5 « 00 t3 2 p H d • 101 Figure 71. Geographic distribution of bluefin release and recapture data shown in Table 27. Cabo has apparently provided first- hand evidence against the existence of the thermal or halothermic barrier to maturing bluefin at the Strait pos- tulated by Roule (1914a), F. de Buen (1931) and others. The sizes of schools were also difficult to determine. Lozano Cabo estimated that one of the largest schools, observed on June 21, 1957, between Algeciras and Tarifa, Spain, occupied about 7,200,000 m\ If the fish were 20 m apart in all dimen- sions, their number would have been 1,922. If the spacing were reduced to 10 m, the number would be 11,163. Lozano Cabo estimated the av- erage speed of the schools at less than 7 knots (13 km/hr). He pointed out that, although the schools ob- served in the Strait were numerous and large, it should be recalled that 1957 was a very good tuna year. Some uncertainties may exist in regard to the actual origin of Sella's (1927, 1929a) hooks and the identity of Lozano Cabo's (1959b) sonar tar- gets, but Rodriguez-Roda (1963, 1964c, 1969a) provided indisputable proof of bluefm tuna migrations from the Atlantic into the Mediterranean. He marked 3 1 2 bluefin tuna, ranging from 60 to 220 cm in length, from catches in traps near Cadiz, Spain, in 1960-1968. Four of these fish were recaptured in the Mediterranean and 15 in Atlantic waters off southern Spain and Portugal and western Mo- rocco (Table 27, Figure 71). The Mediterranean returns are of special interest, in view of the long debate as to whether or not im- portant numbers of bluefin tuna mi- grated through the Strait of Gibraltar. Two of the fish were recaptured at La Linea and Ceuta, Spanish ports on the north and south sides of the Mediterranean outlet of the Strait, respectively, and about 40 nautical miles (64 km) from the release point. The other two, recaptured off Cartagena, Spain, and Palavas, France, had penetrated much more deeply into the Mediterranean, 270 and 675 nautical miles (43 1 and 1 ,080 km) respectively, from the release point. Three of the releases were in May and June, during the "arrival" run when the fish are supposed to be travelling eastward. The fourth, how- ever, was during the "return" run, when they are thought to be travel- ling westward. The recaptures were all in July or August, during the "re- turn" run. Times at large varied from 19 to 23 days, and the weights (jf the fish were 41, 83, 90 and 120 kg (me- dium size range). These returns dem- onstrate conclusively that some blue- fin enter the Mediterranean from the Atlantic. They also suggest that some of these fish return to the Atlantic, since the traps at La Linea and Ceuta, at the extreme western end of the Mediterranean, are designed to fish the "return" (westward) run only. The recovery of two Spanish tags from bluefin tuna caught in a trap near Tripoli, Libya, was reported, but the tags have not been returned, nor have their numbers been ascertained (Rodriguez-Roda 1969a). The Atlantic recaptures (15) were much more numerous than those in the Mediterranean (4), but the prob- ability of recapture of a tagged blue- fin seems to have been much greater in the Ibero-Moroccan Bay than in the western Mediterranean. The fish- eries in the former area took many more bluefin in the period 1 960- 1 967 than those in the latter. Much more tagging is needed to evaluate the mi- grations between the Atlantic and the Mediterranean quantitatively. Much meristic and morphomet- ric data on bluefin tuna from various areas of the eastern Atlantic and the Mediterranean are available, and sev- eral comparisons of the data for fish from different localities have been made (Tiews 1963). Sella (1929a, 1930) noted that he had made several observations on tuna from Spain, Calabria (Italy) and Tripoli (Libya), without finding differences sufficient to justify any racial demarcation. He did not, however, publish these data. Frade (1931) concluded that there were significant differences between the tuna which he had examined on the Portuguese coast and those which Heldt (1927b) had measured in Tu- nisia. Nedelec ( 1 954) compared mor- phological data from bluefin tuna cap- tured in the North Sea, off southern Portugal and off Tunisia. He con- cluded that the North Sea samples were close to those from Portugal, but that both differed considerably 102 from those from Tunisia in many characters. Arico and Genovese (1953) and Genovese (1957, 1958) compared morphological data for spawning and feeding tunas taken off northeastern Sicily with similar data for bluefm taken off Tunisia and southern Portugal and in the North Sea. They concluded that the two Sicilian samples were from a single Tyrrhenian stock which was distinct from the fish from the other three areas and also from smaller samples taken off the Mediterranean coast of France, and off Algeria. Tiews ( 1 963) summarized, com- pared and discussed these data and studies. He found that only the dif- ferences in head length, pectoral fin length and number of finlets support the case for the existence of separate stocks on the European side of the Atlantic. In view of the long and rapid migrations of bluefm demonstrated by tagging experiments, however, he did not believe that the above differ- ences would stand up under critical inspection. He predicted that further studies would prove the existence of a single bluefin population in the east- em North Atlantic. Genetic and biochemical char- acteristics of bluefin tuna from the eastern Atlantic and the Mediterra- nean have also been investigated and compared. Keyvanfar ( 1 962) studied the serology and immunology of bluefm tuna from the Atlantic and Mediterranean coasts of France. In- dividual differences in immunology between Mediterranean samples were found but the samples were too few in number, particularly those from the Atlantic, to furnish any inter- pretation on migatory tendencies. Lee (1965) provided further information on the serology and immunology of the bluefm from the French Mediter- ranean coast and ( 1 968) comparisons of the immunology of bluefin tuna from the eastern Atlantic (30 speci- mens) and the Mediterranean (72 specimens). Lee found individual dif- ferences even within the samples from the Atlantic and those from the Mediterranean. She also found that the percentage of individuals with no antigens and those with one or sev- eral differed in the two lots. She con- cluded that one might consider the presence in the Gulf of Lion (Mediterranean) of certain individu- als which belong to a race or to a population different from that ob- served in the Bay of Biscay. These biometric and biochemi- cal studies provided interesting information, but, in our opinion, they were inconclusive in determining whether the bluefin in the Mediterra- nean and eastern Atlantic constituted a single or separate stocks, or to what degree mixing occurred. In summarizing his extensive studies of the fisheries and migra- tions and other aspects of the biology of the species, Sella (i929a, 1930) proposed the following four phase migratory cycle for bluefin tuna: a. Gathering of bluefin in May-June in relatively limited spawning ar- eas in the southern part of its range (Mediterranean, Ibero-Mo- roccan Bay) b. Post-spawning dispersion, mainly northward, with the maximum extension of their habitat occur- ring in the wannest part of the summer and autumn. c. New reduction in habitat with the coming of winter. This distribu- tion was not well known, because the fish were farther below the surface, but was presumed to be considerably reduced, with ma- jor withdrawal from the more northerly area. d. In March-April, immediately pre- ceding the reproductive period, the tuna made another northward migration which was much more limited than that in summer-au- tumn. They appeared en masse off the north coasts of the Medi- terranean and the Spanish coast of the Bay of Biscay. The subsequent development of fisheries covering nearly all suitable waters of the Atlantic, and modem research, including tagging and track- ing with sonar, have shown how ac- curate Sella's concept was. The only dubious point is his fourth phase, the pre-spawning northward migration. This probably consists of immature fish and the younger mature ones, concerning whose spawning habits little was known then (Sella 1929a), or indeed, even now (Sara 1973). Both authors indicated, however, that these smaller reproducers spawned later than the larger ones. Sara (1964, 1973) has proposed a hypothesis which reconciles many of the seemingly contradictory facts in regard to the relationships between eastern Atlantic and Mediterranean bluefin tuna. He believed that a num- ber, determined by the environmen- tal conditions existing then and in the preceding few weeks, of large (over 100 kg) bluefin tuna gathered in the ibero-Moroccan Bay in late April and early May, "hesitating" (Lenier 1959) there. Then a portion of these, the number again depen- dent on environmental conditions, entered the Mediterranean to spawn, and subsequently returned to the At- lantic. Lozano Cabo (1958) and Aloncle (1964) showed that the average sizes of the bluefin taken in the Atlantic traps, both in Morocco and along the south coasts of Portugal and Spain, varied inversely with the distance of the trap from the Strait of Gibraltar. The largest bluefin (by average weight) have been taken by the traps of Tarifa and Cape Spartel, at the very entrance to the Mediterranean. Actually, more small fish were taken in the more westerly and southerly traps. This was a good indication that the larger bluefin were more apt to enter the Mediterranean than the smaller ones, just as Sara ( 1 973) hy- pothesized. Sara (1973) stated, moreover, that daily catch records collected con- tinuously during the fishing seasons for several years showed that the first catches of the Cape Spartel trap, at the entrance of the Strait, consistently preceded the first of the traps off western Sicily by from seven to nine days. This 900 nautical mile (1,675 km) itinerary would require average speeds between 4.2 and 5.4 knots (7.7- 1 0.0 km/hr). This range does not differ greatly from the migrating speeds observed by Lozano Cabo ( 1 959b), less than seven knots ( 1 3.0 km/hr), and by Rivas (1955, 1976), four knots (7,4 km/hr), or from the average speed of a giant bluefin which migrated from the Bahamas to Nor- way in 50 days (Mather 1969), 3.5 103 knots (6.5 km/hr). These facts refute the argument that no lag between the catches in the respective localities existed, or that it was so short that impossible speeds would have been required for a tuna to have accom- plished the implied migration (Pavesi 1887, 1889, Roule 1925) This infor- mation, combined with Sella's (1929a) findings of Atlantic hooks and lures in tuna caught in the Medi- terranean and the tag returns show- ing similar migrations by Rodriguez- Roda (1969a), strongly supports the old concept that numerous large blue- fin migrate from the Atlantic into the Mediterranean in May and June. The departure of large bluefin from the Mediterranean in the "return" period, July and August, has been indicated by rather small catches in the "re- turn" traps near the eastern end of the Strait (see Section I VC6a). The mod- est size of these catches in compari- son with those of the Ibero-Moroc- can Bay traps may be explained by Sari's (1973) hypothesis in regard to the depths at which the bluefin travel in the respective migratory periods. He maintained that the "arrival" fish swam in the surface layers, follow- ing the inflowing Atlantic current. This is in accord with the findings of Lozano Cabo (1959b) during his so- nar survey of the Strait. Sara (1973) assumed that, on the other hand, the "return" fish swam at deeper levels, following the outflowing Mediter- ranean current. He cited a new fish- ing technique used in the Sicilian Channel as evidence of this (see Sec- tion VIC). Sari believed that this ten- dency to follow the deep current per- sisted until the gonads had shrunk and the feeding urge became pre- dominant. This might occur more readily after the fish had emerged from the Strait into the Atlantic, where the velocity of the Mediterra- nean current diminishes rapidly. The Mediterranean "return" trap at La Linea, on the European side of the Strait, usually caught more large blue- fin than those on the African side. This is in accord with the tendency of the "return" tuna to follow the European coast of the Ibero-Moroc- can Bay closely, as evidenced by the formerly large "return" catches taken there, in contrast to the lack of "re- turn" trap fisheries along its African coast. Sari (1973) believed that the smaller bluefin (less than 100-150 kg) did not participate in this Atlan- tic-Mediterranean migratory pattern. He felt, as did Sella ( 1 929a), that the immature individuals were relatively sedentary. He concluded that the in- dividuals in this size range made longer migrations after reaching ma- turity but that those in the Mediterra- nean spawned there and those in the Atlantic reproduced in that ocean. At some period in their development as their weight approaches 150 kg, groups of these fish tend to join the groups of larger ones in their Atlan- tic-Mediterranean migrations. Sara ( 1 973) speculated that this change in life style is related to some change in the physiology of the animal, most probably the attainment of the full development of the swim bladder. The sizes of the four fish tagged by Rodriguez-Roda (1969a) near Cadiz and recaptured in the western Mediterranean do not fully support Sara's view in regard to the sizes of fish which usually follow this route. One of them weighed 120 kg, but the weights of the other three ranged from 41 to 90 kg. Only one of these smaller fish, however, could have accom- plished this migration during the "ar- rival" period which Sara (1973) was discussing. TTie results indicate that casual migrations through the Strait by bluefin of various sizes may oc- cur at any time, but that the mass periodic movements are carried out mainly by large individuals. One fact not explained by Sara's (1973) theory is the continued productivity of the fisheries for small bluefin off the Atlantic coast of Mo- rocco and in the Bay of Biscay (Sec- tion IVC5). Since there is no hard evidence of extensive spawning of bluefin tuna in the Eastern Atlantic (Section VD2), important recruitment to these fisheries from the Mediterra- nean seems logical. F. de Buen ( 1 925) hypothesized a migration of small bluefin fi^om the Mediterranean to the Atlantic in the autumn of their second year of life when they weighed 4-5 kg. We believe that it is more probable that this migration oc- curs mainly in the autumn of their first year of life, when they weigh about 1 kg. This ties in with the dis- appearance of age bluefin weigh- ing about 1 kg from the Mediterra- nean coast of Spain in October, and the regular appearance of age blue- fin weighing about 1 .5 kg off the Atlantic Moroccan coast in the sec- ond half of November (Section 1VC5). The massive catches of age bluefin in the fall by the "return" traps and purse seiners at the western end of the Mediterranean (Rodriguez- Roda 1964b, 1969d; Crespo and Rey 1976) also suggest such a migration. In 1963, captures around Ceuta in- creased sharply in mid-September and remained high until late Octo- ber, when they declined. c. Migrations and Stocks Within the Mediterranean Much has been written about the migrations and stocks of bluefin tuna in the Mediterranean Sea, but the tag- ging results required to reach defini- tive conclusions are still lacking. De- ductive studies have lead to widely diverging opinions, but a reasonable working hypothesis has emerged. Since many of the pertinent stud- ies have already been discussed in parts a and b of this section, they will be recapitulated briefly here, stress- ing the aspects which are relevant to migrations within the Mediterranean. According to Aristotle's (circa 325 B.C.) theory, the bluefin was essen- tially an Atlantic fish, but spawned in the Black Sea. Therefore it made the circuit of the Mediterranean by passing through it once, en route to the spawning ground, and again to return to the ocean. Little new information was added until Cetti (1777) made two very important discoveries. One was that bluefin spawned near Sardinia. He surmised correctly, moreover, that more of them spawned in the central Mediterranean than in the Black Sea. His other major finding was that me- dium sized bluefin occurred in the Gulf of Sardinia throughout the year. He named these fish "golfitani", a term which is still in common use. He adhered to Aristotle's basic con- cept, however, and described several distinct routes by which he believed that the larger bluefin reached the 104 central Mediterranean. This last find- ing was criticized (Pavesi 1 887), but recent evidence (SarA 1 973) indicates that it was surprisingly accurate. Pavesi (1887, 1889), who was the first to seriously challenge the migratory theory, believed that the Mediterranean (and eastern Atlantic) bluefin spent most of the year in abys- sal waters near their spawning grounds. His theory restricted their migrations to a vertical ascent to su- perficial water and a short horizontal migration to a nearby spawning ground in spring, and short horizon- tal return and a vertical descent into tlie abyss after spawning. Sanzo (1910a) agreed with Pavesi (1887) in regard to the com- plete separation of the Atlantic and Mediterranean bluefin stocks. He had doubts, however, in regard to Pavesi's hypothetical bathymetric migrations. He recommended investigating the darkness or lightness of the colora- tion and other characteristics of the tuna when they first arrived at the traps to determine whether they had come from the abyssal depths or from the surface layers. He believed that spawning was independent of the coastal movements of the tuna, and that other causes of their comings and goings in the Mediterranean should be sought. Bounhiol (1911a, 1911b) also believed that the Mediterranean blue- fin tuna stock was separate from that of the Atlantic. He introduced a new concept, however, that the bluefin in the Mediterranean always moved against the wind-driven surface cur- rents — the "hydrodynamic" theory. This theory has received scant sup- port. Ninni (1922) discussed hypo- thetical populations and migrations of bluefin tuna in the Mediterranean in considerable detail. He believed that the Tyrrhenian bluefin were in- dependent of the Adriatic ones and that both were independent of those of the Aegean Sea, and that the last were at least partly independent of those of the Sea of Marmara. Ninni assumed each group had its own well defined wintering area. He shared the opinion of Pavesi (1887) that the bluefin tuna wintered in deep waters. Ninni believed that the major eastward migration of bluefin tuna must have departed from the winter- ing area between Sardinia and Tuni- sia, passing along the northeastern coast of Tunis then along or offshore from the Tripolitanian coast toward Bengasi. There Ninni's personal ob- servations ended, but he felt that the bluefin avoided the Nile outfiow and turned toward Crete. Ninni thought that there was a wintering area around Crete, and possibly another between Crete and Alexandria. Egypt. Ninni hypothesized that when the fish left the Cretan wintering area they split into two groups. The smaller group went up the western Aegean, passing between Euboea and the mainland of Greece and entering the Gulf of Volos. The larger group fol- lowed the Asiatic coast to the Dardanelles and into the Sea of Marmara and the Bosphorus, always leaving the islands on their left. They passed through the Bosphorus and "lost themselves" in the Black Sea. The passage into the Black Sea be- gan March 15 and lasted into Au- gust, after which the "return" began, but not all of the bluefin went back into the Aegean. A large proportion of them wintered in the Sea of Marmara. Ninni's hypotheses do not lead to an annually repeated migra- tory cycle; in successive years the same fish might winter once between Sardinia and Tunisia, once near Crete, and once in the Sea of Marmara. Such shifts of habitat (or migratory pat- tern) are unusual. Roule (1914a, 1914b. 1917, 1924) believed that the bluefin were pelagic or bathypelagic rather than abyssal, but he felt that their migra- tions were restricted to the particular basins of the Mediterranean which they occupied. He (1924) studied temperature and salinity observations made during the season of spawning assembly (May-June) in 1923 be- tween southern France and Tunisia. He found that the line of maximum thermal increment ran directly from southern France to Tunisia, fitting his (1917) hypothesis that uhen the bluefin were absent from the former area, they were in the latter to spawn. He found that these observations sup- ported his halothermic theory in re- gard to water temperature, but showed no correlation in regard to salinity. Sella (1927) disputed Roule's (1917, 1924, 1926) hypothesis that the tuna which occurred off the south- em coast of France during the re- mainder of the year spawned in May and June near the traps off Sardinia, Sicily and Tunisia. He showed that the tuna caught off the southern coast of France averaged only about 20 kg, whereas those caught in the traps av- eraged from 70 kg to 130 kg. The differences between Roule's and Sella's opinions in regard to the sen- sitivity of the bluefin to the tem- perature and salinity of the water, and the effects of these factors on its distribution and migrations, have been discussed in Section VE2. Scordia (1938) maintained that the bluefin tuna which she studied off eastern Sicily and the west coast of Calabria (southern Italy) were of a distinct and separate stock which she called the Tyrrheno-lonian stock. She observed that these tuna remained mainly in the deep waters of the lower (southern) Tyrrhenian in autumn and winter, sometimes rising to the sur- face near Messina in the fall to feed. They surfaced in the spring with the wanning of the waters to 18°C. When increased warming reduced the den- sity in siiii, from between 1.02700 and 1.02800 to about 1.02500, the tuna moved from the Tyrrhenian through the Strait of Messina into the Ionian Sea, where the density was higher. This occurred toward the end of May and in the first half of June. During June the continued warming reduced the density in the Ionian Sea. The tuna then returned to the deep waters of the Tyrrhenian where the density was suitable for them. They went a few at a time, with the large fish departing first, followed by smaller ones up to September. These movements, which constituted the "arrival" and "return" runs, were based on the sensitive reactions of the fish. At the beginning of their period of sexual maturity they un- derwent a reversal of their thermo- tactic reaction. During the winter they chose a water temperature of about 13.5-I4.5^C. but with the approach of the spawning period, they became 105 thermophiles and sought surface wa- ters of at least 18°C. The migrations were probably provoked by the "transgressions" of Atlantic water, which controlled their period and vol- ume, and were independent of gonad condition (Scordia 1932). The biometric studies of Arico and Genovese (1953) and Genovese (1957, 1958) supported Scordia's hypothesis of a distinct and separate Tyrrhenian-Ionian stock of bluefm tuna. Sella ( 1 929a) maintained that the Mediterranean bluefin could not be separated into autochthonous stocks corresponding to the various basins, as proposed by Roule (1917) and Ninni (1922). In addition to his Atlantic-Mediterranean results, Sella's (1927, 1929a) hook recover- ies indicated that bluefin tuna were constantly moving from one place to another within the Mediterranean (see Heldt 1930a, his unlabeled figure). These findings will be considered in terms of three general areas of ori- gin: the western Mediterranean, cen- tral Mediterranean, and the Bosphorus (Istanbul, Turkey). Hook items of Atlantic origin which were retrieved in the Mediterranean will be considered with those of the west- em Mediterranean, since the fish in which they were found could only have entered the Mediterranean through the Strait of Gibraltar. Hook recoveries indicated exten- sive eastward migrations within the Mediterranean. Numbers in paren- theses following deduced migrations show the number of such migrations revealed, if more than one. Items from Malaga were recovered at Arzeu, Al- geria, and near Tripoli, Libya. One from Arzeu was found off southwest- ern Sardinia. Another, from Philippeville, Algeria, was retrieved at Sidi Daoud, Tunisia. More numer- ous and extensive eastward migra- tions were deduced from recoveries of fish with Atlantic hooks, whose Mediterranean movements must be considered to have begun at Gibraltar. These retrievals occurred off Sardinia (10 + 1 probable), Sicily (5), Gallipoli, Italy southeast of Taranto (1)), Tunisia (4 + 1 probable) and Tripolitania (3). Thus movements from Gibraltar as far east as the heel of the Italian boot on the north coast (longitude IS^E) and nearly to Cape Misurata on the south coast (longi- tude 1 5°E) are indicated. Recoveries of single hooks or lures of central Mediterranean origin suggested that bluefin moved freely between the various fishing centers: Sardinia to Tunisia, Messina to Sardinia, and Palermo. Sicily, to the Aegades Islands. More numerous re- ported findings showed movements from the Strait of Messina to the "re- turn" traps along the southern part of the east coast of Sicily, which cap- tured fish travelling southward along the coast. The only movements to other regions were a westward one from Messina to Arzeu, Algeria, and a longer migration to the east and north, from Sicily or Tunisia to the upper Adriatic near Trieste, Italy. More dramatic findings resulted from six recoveries of very distinc- tive and easily recognized lures used only in the vicinity of Istanbul, Tur- key. Two of these recoveries indi- cated migrations to eastern Tunisia and southwestern Sardinia. Sella ( 1 929a) considered the finding of four lures, each in a different individual caught in the same season and in the only trap operating in the area around Bengazi, Libya, as proof of the ar- rival of several schools coming di- rectly from the Sea of Marmara and the Bosphorus. Despite these impres- sive findings, the tendency to split the Mediterranean bluefin into sepa- rate stocks continued for another 30 years. The taggings of bluefin in the Mediterranean have not been suffi- ciently numerous to answer the ques- tions which have been raised by de- ductive research. Arena and his col- leagues marked 288 bluefin tuna off Sicily and the Aeolian Islands in the years 1963-1968 (Arena and Sara 1967, Arena and Li Greci 1970, Arena 1 97 1 ). Most of these were very small individuals (age 0, lengths 28- 42 cm), but eight were adults weigh- ing 28 to 60 kg (ages 4-5). Returns from four small fish and one adult have been recorded. The recapture of another small fish was reported but the tag was not returned. The returns for small fish were from releases in October 1967. The recaptures occurred after periods at liberty of 65 days or less and within distances of 150 km from the release points. One had passed through the Strait of Messina into the Ionian Sea, but the others had remained in the southeastern comer of the Tyrrhenian Sea. The unconfirmed recovery from a small tuna reportedly occurred within the general release area after a time at large of about 18 months. The adult tuna had made a longer migration, from off Punta Raisi near Palermo at the end of May 1968, to off Castellon, Spain, in mid-October 1969 (Figure 71). Its weight was es- timated as 28-30 kg when released, and reported as 45 kg when recap- tured 17 months later. The returns from small fish sug- gest that their movements are quite limited, but the times at liberty were not sufficient, except in the case of the one unconfirmed recapture, to establish this as a definite tendency. The migration from Sicily to Spain by the larger fish, which was prob- ably age 4 when released and age 5 when recaptured, supports Sella's (1929a) view that the mature bluefin travel freely about the Mediterranean, and refutes the concept of a distinct Tyrrhenian stock (Ninni 1922, Scordia 1938, Arico and Genovese 1953, Genovese 1957. 1958). Obvi- ously, much more tagging in various parts of the Mediterranean is needed to solve the complex problems of the migrations and stock structures in that sea. Sara (1964. 1973) studied the behavior of bluefin tuna in the Mediterranean extensively. He be- lieved that age bluefin made only very limited local movements in the area where they were spawned. As they grew larger, they migrated far- ther, but always within the same ba- sin, until they attained a weight of about 100 kg and joined in the Atlan- tic-Mediterranean migration. The re- sults of tagging in the Mediterranean cast doubt on the hypothesis of a separate Tyrrhenian-Ionian stock (Scordia 1938, Arico and Genovese 1953, Genovese 1957, 1958). Sara (1973) described the mi- gratory and distributional patterns of bluefin in the Mediterranean in terms of the size groups. He used size 106 groupings recommended at the "Re- union sur ie Developpement et la Coordination de Programmes de Recherches sur Ie Thon en Mediterranee, Palermo 22-24/5/ 1967" (Anonymous 1968). These size ranges, which are quite similar to those we have used, were as follows: large spawning tuna of over 150 kg, medium spawning and feeding tuna of 15 to 60-70 kg, very small tuna and small tuna of up to 15 kg which have not yet at- tained their first maturity. Sara believed that most of the large spawners which were observed from the end of April through all of August, were of the same stock, of which a portion, greater or smaller in relation to certain oceanographic con- ditions, entered the Mediterranean to spawn and, having accomplished this, returned to the Atlantic to reconsti- tute their biological reserves (Sec- tion VIC lb). The numbers of tuna "hesitating" in the Ibero-Moroccan Bay and the volume of water enter- ing the Mediterranean determine how many of these tuna enter in a given year. These "arrival" tuna swim in the surface layers, following the branches of the Atlantic currents or its counter currents. These usually follow routes quite similar to the much-criticized ones proposed by Cetti (1777). Displacements of these branches of the currents by the winds, toward or away from the traps, has a critical effect on their catches. These currents were illustrated by Sara (1964, his Figure 1). The main current generally followed the African coast, bringing the tuna within range of the traps of Tunisia and Tripolitania. A west-flowing counter current brought the "arrival" tuna into the traps along the north coast of Sicily from the east, rather than from the usual westerly direc- tion. According to Sara (1973), the "return" tuna swim in deeper layers, following the westward flow of the Mediterranean water. This current flows southward and westward along the east and south costs of Sicily, where "return" traps were set until recent years. Sara (1973) found evi- dence of this in the new night fishery practiced in waters from south to southwest of the western end of Sic- ily. Dead bait was thrown into the water (chumming) to attract the tuna to the surface. There they fed on small live fish which had previously been drawn to the boat by powerful lights. The tuna were then vulnerable to cap- ture by seining. These catches took place from late July until mid Sep- tember. Some large tuna remain in the Mediterranean through the fall and winter in areas of rich and available food, such as the islands of the Tuscan Archipelago, the area between the Aeolian Islands and the Strait of Messina, the Dardanelles, and Port de Bouc in the Gulf of Lion. The numbers of these fish are in- consequential in comparison with the great mass of migrating fish which occur in the period of reproduction. The schools of large "Atlantic" tuna remain separate, according to size offish, during the period of matu- ration. During the time of actual spawning, however, they mix tem- porarily with each other, and with younger Mediterranean fish of ages 4-7. The latter at other times remain altogether aloof from the larger fish, and, even during the reproductive processes, occupy other layers and areas when constrained together with them. Medium and small fish weigh- ing from 20 to 50 kg, with fringes of larger and smaller individuals, are seen and captured along the coasts of the Mediterranean countries through- out the year (Sar^ 1973). These fish school very strictly according to size, They are seen closer to the coast in late summer and autunm, often on the surface in constant and furious pursuit of prey. They are present all year in certain locations near the coast where they can be observed and fished, even in bad weather, and the currents and geomorpholog) of the bottom are favorable, as at the Straits of Bonifacio, in the Adriatic, at the Aeolian Islands, at the Strait of Messina, at the Kerkennahs, and at the Dardanelles. These fish have a different mi- gratory pattern from the large ones and make limited spawning and tro- phic (feeding) migrations, within the limits of each basin. Apparently it was their contemporaneous presence in several localities, and their habit of remaining off this coast or that one, that, years ago, led to the con- cept of various populations differing at the racial level. In the spawning period, they approach the coasts, ag- gregating with the schools of large Atlantic tuna with which they can mix only during the limited period of reproduction. Sara (1973) considers the small and very small tuna next. In the months from August through all of November, almost all of the Medi- terranean Basin is occupied by great schools of young of the year. Re- maining for a few months along any coast, one can see them grow day by day until in November they attain about 2 kg. After this, probably more because of bad weather than because of their departure, they are rarely cap- tured. With the return of good weather, ihey are found again in the late winter months in almost the same places where they had been seen ear- lier, but now weigh 4-5 kg. The young tuna stay in the same areas, making small movements in response to meteoro-hydrological conditions and in search of food. They remain sed- entary until they attain their first sexual maturity, when they begin their first migrations about the Medi- terranean. Sara's (1973) concept of the mi- gratory and distributional patterns of bluefin tuna in the Mediterranean is very similar to Sella's (1929a) in re- gard to the large bluefin, which both consider to be essentially migratory, and the immature fish, which both consider to be relatively sedentary. They differ somewhat concerning the behavior of the intermediate (me- dium) fish. Sella (1929a) felt that these were constantly moving, and could not be separated into groups in the various basins, whereas Sara (1973) believed that their aggrega- tions tended to remain stationary for considerable periods, and migrated only within the basin which they oc- cupied. Both scientists agreed that knowledge of the spawning behavior of these intermediate bluefin was in- adequate. 107 The few details available on the trap fisheries of Tripolitania, and the ephemeral one in Cyrenaica (Bengasi) may provide clues in re- gard to this important subject. Both Sella (1929a) and Sara (1973) cited the later fishing of the Libyan traps, compared to those in Tunisia, as evi- dence of the eastward "arrival" run of the bluefin from the Atlantic. Sella (1929a) noted, however, that the fish taken in Tripolitanian and Cyrenaican traps were medium and small. In a later work. Sella (1932a) stated that the fish taken by the El Mongar trap near Bengasi, Cyrenaica, although mature, were too small to justify its classification as an arrival trap. Anonymous (1929b) described the fish which it caught in 1928 as very small ("picolissimi"). These data in- dicate that the Libyan traps may have been harvesting the intermediate (Mediterranean) mature fish rather than the large (Atlantic) migratory spawning group. Another unsolved question con- cerns the migration between the Mediterranean and the Black Sea. Hovasse (1927), Akyuz and Artiiz (1 957) and lyigUngor ( 1 957) all cited passages of maturing fish through the Bosporus into the Black Sea, and of spent fish in the opposite direction (Section VDl). They differed, how- ever, in regard to the dates of these runs. Hovasse (1927) and lyigungor (1957) set the northward migration in March and April, and the begin- ning of the southward one in July. Akyxiz and Artiiz (1957) postu lated a longer northward passage, starting in April, peaking in July, and extending into September, and a "return" run occurring from late October into No- vember. Sella ( 1 929a), on the other hand, cited what he considered to be clear evidence of a massive migration of tuna from the vicinity of Istanbul to spawning grounds off Cyrenaica. Four lures, of special types used only in the Bosporus and its approaches, were retrieved from the few hundred tuna caught during the 1929 season in the EI Mongar trap, the only one set in Cyrenaica. Such a migration would be consistent with Sella's (1929a) hypothesis of spring spawn- ing concentrations in the southern parts of the bluefin tuna's range (Sec- tion VIClb). These apparently inconsistent findings might be reconciled if it could be shown that the larger fish in the Sea of Marmara entered the Black Sea to spawn, whereas the smaller mature individuals reproduced off the African coast. Another probable spawning area for the intermediate bluefin is in the Adriatic Sea, where these fish are much more abundant than giants (Section 1VC6, Section VDI). This would be more in keeping with Sara's (1973) concept of limited spawning migrations, within basins, for the smaller mature fish. The configuration of the Medi- terranean and its approaches is such that many important problems could be solved quickly and definitively by tagging, or by tracking with sonic equipment. d. Migrations and Stocks in the Eastern Atlantic i. Large and Medium Fish In addition to the four fish re- captured in the Mediterranean (Sec- tion VIC I b). 1 5 of the 3 1 2 fi.sh tagged in the years 1960-1967 near Cadiz have been recaptured in Atlantic wa- ters (Rodriguez-Roda 1969a). These recaptures provided information on several migrations within the area of the Ibero-Moroccan Bay trap fisher- ies, and one to a more distant locality (Figure 71). Since the periodic migratory pas- sages through the trap fisheries ex- tend over only about four months, short-term tag returns are required to study them in detail. All of the six recoveries from 140 fish tagged dur- ing the "arrival" period occurred within 85 days of the releases. As noted in part b of this section, three of these recoveries look place in the Mediterranean, representing a con- tinuation of the expected easterly movement. One of the other recover- ies was irrelevant, since the fish was recaptured almost immediately b>' the trap from which it had been released. The other two recoveries, however, indicated rather rapid "backward" (westward) migrations of 79 miles (146 km) in one day and 100 miles (185 km) in two days, respectively. into Portuguese waters. These fish were 170 cm and 133 cm long, re- spectively. Seven of the 13 recap- tures from the 172 fish released dur- ing the "return" run occurred after from three to 23 days at large. One recapture, as noted in part b of this section, had "backed" to the east- ward, well into the Mediterranean. Another, recaptured where it had been released tiiree days previously, pro- vided no information. The remaining five made rather rapid migrations of 90 to 100 nautical miles (167-185 km) in from four to 1 1 days, travel- ling in the expected direction into, or approaching, the waters south of Por- tugal. Three of these fish weighed from 70 to 90 kg, one weighed 270 kg, and the weight of the remaining fish was not reported. These short term recoveries suggested that fish tagged during the "arrival" period were more apt to deviate from their expected migratory pattern than those marked in the "return" period. This is rather surprising, since the "ar- rival" fish are believed to be concen- trating to spawn, whereas the "re- turn" fish are thought to be dispers- ing to feed. The approximate times at large for the remaining six recaptures of fish marked in the "return" period were 4 months for one fish, 10 months for three fish and 22 months for the other two fish. The longest Atlantic migration revealed by this set of re- leases near Cadiz, about 370 nautical miles (685 km), was from off Barbate, Spain, in August to off the Moroccan coast between Essaouira and Agadir four months later (Figure 71). This movement supported Aloncie's ( 1 964) hypothesis of seasonal migra- tions of bluefin tuna between waters off southern Morocco and the Ibero- Moroccan Bay. This fish, however, weighed 175 kg, considerably more than the group which Aloncle was discussing. All of the long-term (10 months or 22 months at large) recaptures of fish marked during the "return" run were in traps fishing the "arrival" passage Three recaptures occurred near Earache, Morocco, about 40 to 60 nautical miles (74-1 12 km) south of the release point, one off Portugal and another at Sancti Petri, about 100 108 nautical miles (185 km) and about 20 miles (37 km) west of the release point, respectively. These long-term recoveries suggested that the Moroc- can and Iberian trap fisheries in the Atlantic harvested the same stock. Comparison of the return rates and times at liberty for fish tagged from the "return" run (7.6%, five re- captures after 302-683 days at lib- erty) with those for fish marked dur- ing the "arrival" run (4.3%, no re- captures after more than 85 days at liberty) suggests that fish tagged fi^om the "return" run were more viable than those released from the "arrival" run. Giant and medium-sized bluefin occurred in Scandinavian and North Sea waters during the summer and fall. It was generally believed that these came, at least in part, from the "return" run which was fished by the traps off southern Spain and Portu- gal in July and August (Sella 1929a, Sara 1964). Le Gall (1929) projected a northward migration route west of Ireland and north of Scotland into the North Sea and the Scandinavian waters, on the basis of observations of tuna from fishing boats along the way. Sella (1929a) identified a hook found in a bluefin tuna caught in Oslo Fjord, Norway, as of the type used in the Bay of Biscay. This find- ing supported the concept of a north- ward migration of bluefin tuna from Iberian waters to those off Scandinavia. Le Gall (1929) also noted indications of a southward mi- gration passing north and west of the British Isles in the fall. Le Gall ( 1 927, 1 929) noted that the presence of blue- fin in the North Sea was controlled by influxes of Atlantic or Atlantic slope waters with salinities of about 35 o/oo. These waters entered the North Sea from the north. Records of bluefin in the English Channel were scarce, although fish occurred fre- quently off its western entrance, and off the western end of Cornwall. Le Gall ( 1 927) noted that these fish were in the feeding phase, and speculated that they had spawned in the Ibero- Moroccan Bay. The accuracy of this assumption was proved by the tagging experi- ments of Hamre (1965). Hamre tagged 242 medium and large (150- 250 cm long) bluefin tuna in August and September 1957-1962 near Bergen, Norway. Six of these fish have been recaptured near Cadiz, Spain, and 26 in Scandinavian wa- ters (Figure 72 ) (Hamre 1965 and personal communication). Four of the Spanish recaptures occurred after less than 10 months at liberty ("direct" migrations). The other two took place 23 and 45 months after release, respectively. If we exclude nine of the local returns which were recov- ered in the season of their release, and thus had no opportunity to emi- grate, 26% of the returns were from Spain, as against 74% from local waters. Bluefin were tagged off Nor- way in each of the years 1957-1962, but the tags used in 1957 were evi- dently defective and produced no re- turns at all, and only 13 fish were marked in 1962. One more of the fish tagged off Norway in each of the other years was subsequently recap- tured off Spain. These results sug- gest strongly that substantial num- bers of the bluefin tuna which oc- curred off Norway in summer and early fall also occurred off southern Spain in spring and early summer, presumably there to spawn. Studies of the size composition of the land- ings in Norway and southern Spain (Hamre et al. 1968), however, indi- cated that the fisheries in these two areas did not harvest the same stock in every year. In 1955 through I960, the size composition of the Norwe- gian and Spanish catches were quite similar, but in 1961 through 1964, they were quite different. In 1965 they were similar again. Hamre et al. (1968) attributed these changes to a subdivision of the northeast Atlantic tuna population into two contingents with different migratory habits in 1961 through 1964. Although it has not been dem- onstrated by tag returns, Hamre agreed with Le Gall (1927, 1929) that the bluefin entered the North Sea and Scandinavian waters from west and north of the British Isles, rather than through the English Channel (Figure 73). After the disappearance of me- dium-sized bluefin from the Norwe- gian fishery in 1963, the northward migration of the giant bluefin along the Norwegian coast ceased and most of these fish followed the southward migratory route previously used by medium-sized fish (Hamre 1965). Tiews (1964) and Hamre (1965) agreed, on the basis of the locations and periods of fishing and the sizes of fish taken, that the large bluefin which had arrived in July and trav- elled northward along the Norwegian coast from Bergen eventually left that coast and migrated southward into the North Sea. There they were vul- nerable to the German and Danish fisheries (Section IVC5 b and c). When these large fish reversed their migration route in 1963, predomi- nantly following the Norwegian coast southward rather than northward, the North Sea fisheries disappeared (Hamre 1971, Tiews 1975). The local tag returns (Hamre 1965, personal communication) con- firmed migratory patterns (Figure 73 ) in Norwegian waters which Hamre (1962) had also deduced from stud- ies of the size composition of catches. The bluefin tuna reached the coast at about 62°N latitude. The larger indi- viduals, which arrived in July, mi- grated northward and the smaller ones, which arrived in August, mi- grated southward. One of the latter which had been tagged near Bergen was recaptured off the island of Anholt in the Kattegat. In addition to the nine returns in the release season, nine tags were recovered after about one year at liberty, three after about two years, as well as single ones from fish which had been at large for about three, five, six, seven and eight years. In recent years, most of the blue- fin tuna in Norwegian waters, all of which have been very large, have left there early in the fall. Prior to 1963, when medium fish were abun- dant there, many of these fish did not depart until late autunm. The ensu- ing migration of the large bluefin apparently takes them north and west of the British Isles to their wintering area This probably lies between lati- tudes 25°N and 40°N and between the African and European coasts and longitude 35°W (Shingu et al. 1975, Fisheries Agency of Japan 1976). The size distribution of the fish in the area is not known. Aloncle (1964), however, reported that small to me- 109 dium fish occurred between the Ca- nary Islands and Morocco in winter. Many medium and some large bluetln are captured in the Bay of Biscay from mid- July to mid-August, or early September (Creac'h 1952, Le Gall 1954, Bard et al. 1973, Dao and Bessineton 1974, Cort 1975, 1977). Some of the smaller fish may spawn on their arrival in the Bay (Cort 1977), but this is primarily a feeding concentration. Where these bluefin go after they leave the Bay of Biscay is uncertain. Cort ( 1 976) con- jectured that those in age groups 6 to 9 continued their northward migra- tion to Norwegian waters, whereas those in age groups 4 and 5 followed a different, probably more southerly, route. The migration of age groups 6-9 to Scandinavian waters may well have taken place in years prior to 1963, but fish of these ages have not been important in the Norwegian catches since then (Hamre 1971). The European landings of me- dium bluefin in the Bay of Biscay have increased markedly since 1973, perhaps in consequence of substan- tial fishing effort by Japanese longline vessels in the area (Cort and Cendrero 1975, Cort 1976, Fisheries Agency of Japan 1976, Shingu and Hisada 1976). It is unfortunate that adequate size composition data for Bay of Bis- cay landings in the 1960s are not available. It would have been of in- terest to note whether the departure of the medium fish from the Norwe- gian fishery in 1963 through failure of recruitment (Hamre 1971) was re- flected by any increase in their avail- ability in the Bay of Biscay. The medium and large bluefin which leave the Bay of Biscay in August or early September presum- ably reach the wintering area de- scribed above in late fall or early winter, but their whereabouts in the meantime is unknown. Possibly some of them join the concentration off the central Moroccan coast, along with others which may have gone there directly after spawning in the Ibero- Moroccan Bay (Aloncle 1964). In April the medium and large bluefin begin to migrate from their wintering area to their spawning grounds. Most of the kirgc t1sh and some of the older mediums partici- pate in the "arrival" run, assembling in the Ibero-Moroccan Bay in May and June. One contingent of them, mainly the individuals weighing less than 150 kg, presumably spawns there, along with a portion of the larger fish. Apparently many of the latter, however, enter the Mediterra- nean, spawn there in June or early July, and make a "return" run into the Atlantic in July and August. Most of the larger fish which have re- mained in the Ibero-Moroccan Bay also join this "return" run. Some of the smaller individuals may spawn off the Moroccan coast \\b'\\e en route to the Ba\' of Biscay (Aloncle 1964). Others may spawn after arriving there (Con 1 977), thus actually spawning in their feeding area. Most of the large "return" fish migrate from the Ibero-Moroccan Bay around the British Isles to their summer feeding area in Norwegian waters. Many of the smaller ones, including most of those which have spawned in that Bay, proceed to the Bay of Biscay, but others may join the concentration off the central Mo- roccan coast. This gathering is occa- sioned when the impingement on the northern coast of Morocco of a tongue of warm (over 21°C) tropical water Figure 72. Geographic distribution of bluefin tag release and recapture data from Norwegian tagging studies and hook recovery data in the eastem Atlantic. 110 from offshore inhibits the further progress of some groups of small and medium tuna travelling northward from the wintering area. These fish eventually find themselves in the cooler upwelling waters off the cen- tral Moroccan coast, but completely surrounded by tropical water. Most of them stay in this cool water rather than continuing northward through the warmer water (Aloncle 1964). The return to summer feeding areas completes the migratory cycle of the medium and large bluefin of the eastern Atlantic. The hypotheti- cal three phase migration is compli- cated by the entrapment of some me- dium fish off central Morocco, and the entry of larger fish into the Medi- terranean to spawn. ii. Small Fish Thirty-four small bluefin tuna have been tagged in the Bay of Biscay and off Portugal since 1960 (Aloncle 1973). Three of these were recap- tured locally, and two off the north- eastern United States. One of the lo- cal returns was from a small bluefin tagged off Cape St. Vincent (south- western comer of Portugal) in early June and recaptured in the Bay of Biscay 83 days later (Figure 74). This migration supports Aloncle's (1964) hypothesis that small bluefin migrate seasonally from the Ibero- Moroccan Bay in spring to the Bay of Biscay in summer. The transatlan- tic returns will be discussed in part 3 of this section. The results of tagging off the Atlantic coast of Morocco (Lamboeuf 1975) also support Aloncle's (1964) theory. During 1972 and 1973, 63 bluefin, 41-73 cm long, were tagged and released there: 1 5 in June, one in July, 25 in August and 2 1 in Novem- ber. Eleven returns resulted, includ- ing five from Moroccan waters and six from other areas (Figure 74). During 1973 two fish, of which one had been marked in June and the other in August (at ages of approxi- mately 1 1 and 14 months, respec- tively), were recaptured off the south- western coast of Portugal in October and September, respectively. Four others, of which three were tagged in August 1972 (at ages of about 13 months) and one in November 1972, (age about 17 months) were recap- tured in the Bay of Biscay in June- August 1973. These important results indicate that both the minor fall fish- ery off the west coast of Portugal and the much more important one which takes place in the Bay of Biscay dur- ing the warm season draw recruits from off the Atlantic coast of Mo- rocco. These few returns suggest that fish recruit to the former fishery at age 1 and to the latter at age 2. These ages of recruitment are in accord with the age groups which have been pre- dominant in the landings of ihe re- spective fisheries. The tendency for the maximum catches of "cachorretas" (age 1 hiuefin) in the Portuguese trap fishery to have oc- curred in late August suggests that the indicated migration from off Mo- rocco in summer to off western Por- tugal in the fall may sometimes have passed along the south coast of Por- tugal en route. Local returns showed two south- westward movements from the Larache-Casablanca area toward Cap Blanc in August-September 1972 and two northward migrations from off Agadir to off Essaouira in Novem- ber-December 1972. Lamboeuf (1975), considering these results and observations of the movement of the fishery and the sizes of fish caught, formulated the fol- lowing hypothesis: Bluefin spawned off Morocco by the adult "arrival" tuna in spring or earlv summer remain offshore in fa- vorable waters (18°C-22"C). They occasionally approach the coast when hydrological conditions allow this. Having attained 1 year of age and 50 to 60 cm in length, they frequent the coast from June through August, moving from north to south during this period. They then disappear, hav- ing been driven offshore by the cold waters in the Safi-Essaouira region, 6S* 6}« 55« i1° Figure 73. Possible paths of bluefin tuna entering the North Sea and Scandinavian waters from west and north of the British Isles, rather than through the English Channel. Ill Figure 74. Geographic distribution of bluefin tag release and recapture data from Moroccan and Bay of Biscay tagging studies. only to reappear in the Bay of Agadir in November and December. These bluefin which have attained a mean length of 67 cm then undertake a longer migration, passing along the Portuguese coast and penetrating deeply into the Bay of Biscay. In certain years, as in 1 973, when the cold waters of the central sector extended far offshore, the tuna may have migrated earlier in the season, bypassing the Agadir area. These results of tagging off Por- tugal and Morocco and Lamboeuf s (1975) hypothesis explain Fumestin and Dardignac's (1962) observation that young bluefin become scarce along the Moroccan coast in Novem- ber, when they have attained a length of 64 cm, and confirm Aloncle's (1964) hypothetical migration of young bluefin from Moroccan wa- ters to the Bay of Biscay. After feeding in the Bay of Biscay through the remainder of the warm season, the age 2 fish depart, along with fish a year or two older, mainly in October. Their next desti- nation is their wintering area, which is little known. There is a wintering zone between the Canary Islands and Morocco, but the total winter habitat is probably much larger, with the fish spreading over greater areas as they attain older ages. Age 3 and 4, and older, bluefin move northward again in the spring; probably the majority of them even- tually enter the Bay of Biscay. Many of the older individuals in this group are mature. Some of these may spawn during this northward migration, and others may split off from it and spawn with the larger fish in the Ibero-Mo- roccan Bay. During late summer and fall, considerable aggregations of small and medium bluefin occur off the central Moroccan coast, as noted previously. Some of these fish may have been trapped there b> a warm water mass during their northward movement and others may have mi- grated there after spawning in the Ibero-Moroccan Bay. Most of the small bluefin which have spawned there, however, probably proceed to the Bay of Biscay, entering it during the last half of July or early August. Thus the hypothetical two-part mi- gratory pattern of this group may be interrupted by local hydrological con- ditions, and the mature individuals are inclined toward the three-part mi- gratory pattern proposed for the older age groups. iii. Very Small Fish Eastern Atlantic bluefin (apply- ing this term to individuals which are recruited to stocks in the eastern At- lantic as juveniles) are hatched mainly in the Mediterranean Sea and/or the Ibero-Moroccan Bay. Those born in the former area, however, would have to pass through the latter to reach their nursery areas. Supplementary hatching may extend to the area be- tween the Canary Islands and Mo- rocco, and also into the Bay of Biscay. The principal hatching occurs from about the middle of June to about the middle of July, with reproduction by the smaller individuals occurring up to the end of July or later. By mid- November, when they are about 40 cm long, the bulk of the new-born tuna have converged from their vari- ous bii1h places to concentrate off the Atlantic coast of Morocco, where they complete their first year of life in the following June or July. 2. Western Atlantic a. Introduction Tag returns, data on the fisher- ies, and knowledge of oceanographic conditions must be used in combina- tion to establish hypothetical migra- tion patterns for bluefin tuna. Since 1954, about 2,760 giant, 490 me- dium, and 14,630 small bluefin have been tagged in the western Atlantic and 63, 9, and 3,000 returns, respec- tively, have resulted. These returns have not only provided important in- formation on movements within the area, but have also revealed surpris- ing migrations to distant regions. Extensive data on the coastal and oce- an ic fisheries are available (see Sec- tions IVC2 and 1VC3) to comple- ment the pattern disclosed by tag- ging. Also, comprehensive works on the oceanography of the Atlantic, with details of its circulatory system, have been published (Sverdrup et al. 1942). Those concerning the Gulf Stream system (Fuglister and Worthington 1951, Von Arx et al. 1955) are of special interest in rela- tion to the distribution and migra- 112 tions of bluefin tuna in the north At- lantic (Sella 1 93 1 , Rivas 1 955, Squire 1963). Since the economic impor- tance of the species in the western Atlantic has been slight until recent years, the history of research on its migrations there is brief in compari- son with that in the eastern Atlantic and the Mediterranean. b. Large Fish Few deductions about the mi- gration of large bluefm tuna in the western Atlantic have been published. The summer and early fall occur- rences of bluefin tuna off northeast- em North America (off the United States coast from New Jersey to Maine, and off Nova Scotia and New Brunswick in Canada) have long been known (Farrington 1939). A periodic northward migration of "giant" blue- fin through the Straits of Florida (mainly along the northwestern edge of the Great Bahamas Bank near Cat Cay and Bimini) was discovered by sport fishermen in the 1930s and it was soon concluded that these fish were among those which subse- quently occurred off northeastern North America in summer and early fall (Farrington 1939, Mowbray 1949). At the Oceanic Fisheries Con- ference held at the Bermuda Biologi- cal Station for Research, May 28-3 1 , 1951, (transcript of tape recordings available at Woods Hole Oceano- graphic Institution) Schuck and Mather presented their views on sev- eral aspects of the biology of bluefin tuna in the western North Atlantic. They described the behavior of the giant bluefin tuna in the Straits of Florida as observed during flights made with L. R. Rivas in U.S. Coast Guard aircraft. These observations confirmed that the visible schools were virtually all travelling north- ward, as reported by the sport fisher- men. Schuck and Mather also showed that the times of the passage of the giant tuna through the Straits of Florida and their arrival in New En- gland waters were consistent with a migration between the two areas. Fi- nally, they presented preliminary re- sults of biometric studies which showed that the fish taken in the Straits ofFlorida were similar to those taken in New England waters, except in regard to body condition. The fish taken off New England in the begin- ning of the season were similar in body proportions to those taken in the Straits of Florida. As the season advanced, however, the fish off New England fed heavily and gained weight rapidly. Therefore tiie fish taken there in August, September and October became progressively heavier in proportion to their length than those taken in the Straits of Florida. The ratios of their girth and depth to their length consequently also became greater than the corre- sponding ratios for fish taken in the southern area. Rivas (1955), in a similar study, compared the morphological charac- teristics of samples of bluefin tuna of similar sizes taken in the Straits of Florida in May 1 952- 1954(19 speci- mens) and at Wedgeport, Nova Scotia, in October 1952 and August 1953 (nine specimens). He found no significant differences between these samples, except those attributable to the normal seasonal change in body condition experienced by bluefin tuna. These fish lost considerable weight during the spawning process and related migrations, then fed so heavily during the summer and early fall that their length-weight ratio was at a minimum in September-Octo- ber. The lean condition of the fish taken in the early part of each season at Wedgeport was consistent with their having come from the Straits of Florida. In addition, Rivas showed that the times of arrival of giant blue- fin tuna off Cape Cod, Massachu- setts, and Wedgeport were consistent with the time of departure and esti- mated average speed (3.5 knots, or 6.5 km/hr) and direction of travel of the individuals passing the Bahamas. Thus he presented a case, which he considered strong, but not conclu- sive, for a northward migration of giant bluefin from the Straits of Florida in May-June to New England and Nova Scotia waters in summer or early fall. Exploratory and commercial longline fishing has provided impor- tant information on the temporal and areal distribution of bluefin in the oceanic waters of the v\estern Atlan- tic since 1956 (see Sections IVC2 and IVC3). The records of occur- rences during the cold season (No- vember-April), when their where- abouts had previously been virtually unknown, were especially valuable. This information has been used in conjunction with tagging results in developing new hypotheses c the migrations of the species (Mather 1969, 1974, Mather etal. 1974). Continuous tagging of western Atlantic bluefin began in 1954. Since then, about 1 , 100 giant fish have been marked off the northwestern Baha- mas (Straits of Florida) and 15 of these fish have been recaptured over a very extensive area — four off northeastern North America, two off eastern South America, and nine off Norway (Table 28) (Mather 1962, 1969; FAO 1972, Mason et al. 1977). Cooperating sport fishermen provided the effective tagging in this effort. A few fish have been marked during exploratory fishing cruises in this general area, but none of the tags have been returned. The four returns off North America were from ofTMary land to off the Nova Scotian banks (Figure 76); only one was from an area known as a tuna fishing ground. This return was from a giant released in early June 1973, in the Bimini-Cat Cay area (north- western Bahamas) and recaptured about 18 nautical miles (33 km) northeast of Gloucester, Massa- chusetts, a well-known center for small boat tuna fishing, in early July 1974. Since it had been at liberty for 13 months, this was not a "direct" migration. If it is regarded as a sec- ond retracing of an annual migratory cycle, however, the fish would have passed through the Straits ofFlorida again in May or June 1974, and its ensuing migration would have fitted the proposed pattern. The other three recaptures were incidental or acci- dental catches by vessels fishing for other species. While they could be interpreted as representing incom- plete migrations toward the New En- gland or Canadian feeding areas, there is an element of doubt in each case. One return was from a fish re- captured about 48 nautical miles (89 km) east-southeast of Ocean City, Maryland, in April 1973, 23 months after it had been released off the 113 northwestern Bahamas in May 1971. The inshore location of its recapture suggests that it might have eventu- ally entered New England or Cana- dian coastal waters. Its recapture date, April 4, however, indicates that, if it had come from the Straits of Florida, it must have left there in March, well ahead of the usual May-June pas- sage. The other two recaptures oc- curred in offshore waters, from 40 to 65 nautical miles (74 to 120 km) out- side the 2,000 m contour. One fish, which was recaptured 135 nautical miles (250 km) southeast of Nan- tucket Island, Massachusetts, in June 1970 had been released four years previously, in May 1966, in the Cat Cay-Bimini area. The fourth tuna had been released in the same locality in early June 1970, and was recaptured just 30 days later 100 nautical miles (185 km) southeast of Sable Island, off Nova Scotia. Thus it had trav- elled at least 1,500 nautical miles (2,800 km) in a maximum of 30 days, requiring a minimum average speed of 2.1 knots (3.9 km per hr). It is uncertain whether these fish, had their journeys not been terminated, would have proceeded to the feeding grounds off the American coasts, completing the hypothetical migra- tion, or continued across the ocean to Norwegian waters, as so many other bluefin released off the Bahamas have done. It is widely believed, however, that these animals have a strong ten- dency to follow favoring currents when migrating (Sella 1929a, Rivas 1955, Lozano Cabo 1958, 1959b; Sara 1964, 1973; Rodewald 1967). Since both of these fish were recap- tured far north of the Gulf Stream, which would have favored a migra- tion to Norway, it seems more prob- able that they would have proceeded to the nearby American feeding grounds than that they would have wandered out into the less produc- tive waters of the open ocean. Two giant bluefin tagged off the northwestern Bahamas have been re- captured in the South Atlantic (Table 28 , Figure 75) (Mather 1974). One fish was marked in May 1963 and recaptured southeast of Recife, Bra- zil, in March 1965. The other fish, released in June 1969, was recap- tured off Argentina in February 1973. This migration of at least 6,600 nau- tical miles (12,250 km) is the longest ever recorded for an Atlantic bluefin. These surprising recoveries pro- vided the first clues to the relation- ships between the bluefin tuna stocks of the North and South Atlantic. Catch records of the Japanese longline fishery (Fisheries Agency of Japan 1965, 1966, 1967a, 1967b. 1968, 1969, 1970, 1971, 1972, 1973) show that, in some years, an area of intensive bluefin fishing develops in March and April off easternmost Bra- zil, moves northwestward toward the Bahamas and then northward to off Cape Hatteras in May and June. In Jul), the longline catches of large fish diminishes. The capture off Recife of a giant bluefin which had been tagged off the Bahamas in- creases the probability that a group of large fish does actually migrate from off the Bahamas in May-June to off Cape Hatteras by early July, thence to various summering areas farther north (Figure 75), and south- ward in the fall to wintering areas around the Antilles and off South America. It had been generally be- lieved (Wise and Davis 1973) that bluefin tuna concentrations north of Table 28. Releases and returns for giant bluefin tuna (over 120 kg) tagged off the Bahamas by year of release, months at large, and area* of recapture. Also shown is the estimated percentage of German North Sea bluefin tuna catches in late season (September 15 - October 3 1 ) recruited from the western Atlantic (Tiews 1964). Year Releases Returns by Month at Large % 0-5.9 6-17.9 18-29.9 30-41.9 42-53.9 78 - 89.9 1954 21 8 1955 14 4 1956 41 6 1957 1958 7 1959 25 1960 13 2N 11 1961 34 2N 33 1962 45 1 N 2 1963 147 1 B 1964 41 1965 54 1966 105 1 A 1967 82 1 N 1968 57 1969 50 1 B 1 N 1970 182 1 A — 1971 49 1 A — 1972 32 1 N — 1973 47 lA, 1 N — 1974 31 — — — 1975 18 — — — — 1976 5 - — — — Areas: A = Northeastern North America, B = Braz il and Argentina, N = Norway. 114 Figure 75. Geographic distribution of bluefin tag release and recapture data illustrating transatlantic and transequatorial migrations. latitude 20°S consisted mainly of At- lantic bluefin tima T. thynnus thymms. and that those south of 20°S con- sisted mainly of southern bluefin tuna, T. maccoyii. The recapture off Argentina of a giant bluefin which had been tagged off the Bahamas in- dicates that the already tremendous migratory circuit suggested above for Atlantic bluefin occasionally extends much farther south, well into the habi- tat of tlie southern bluefin. nie nine recoveries in Norwe- gian waters Irom the releases in the Straits of Florida will be discussed in part 3 of this section. Ciiant bluefin tuna have also been tagged successlully during the sum- 115 Table 29. Releases and returns for giant bluefin tuna (over 1 20 kg) tuna tagged in St. Margaret's Bay, Nova Scotia, by year of release, months at large and area* c>f recapture. Year Releases Returns by Months at Large Total 0-5.9 6-17.9 18-29.9 30-41.9 42-53.9 54-65.9 1963 18 1964 6 1965 52 2L IL 3L 1966 71 2N 2N 1967 193 5L,1W IN 7LWN 1968 1969 15 IL IG 2LG 1970 3 1971 45 IN 10 IG 3NG 1972 12 1973 1974 8 1975 148 3G IL 4GL 1976 12 IG IG ♦Areas: L = Local, G = Gulf of St. Lawrence, N = New England, W Wedgeport, Nova Scotia Gulf of St Lawrence had been at liberty for 49, 60 and 63 months. These six migrations were therefore ■ indirect ' The recaptures off New England occurred in June (1 recap- ture) and August (3 recaptures); the one at Wedgeport occurred in Sep- tember. Those in the Gulf of St. Lawrence occurred in July (1 recap- ture), August (2 recaptures), Septem- ber (3 recaptures) and October (1 recapture) The two local recaptures after a winter at liberty took place in August and October Sport fishermen cooperating with United States and Canadian pro- grams have tagged 962 giant bluefm in eastern Newfoundland waters, and of these nine tags have been returned (Table 31) Four of these were re- covered locally, but five were re- turned from other areas (Figure 77) The longest migrations recorded were from Notre Dame Bay, Newfound- land, in August 1970 and 1971, to Cape Cod Bay, Massachusetts, in July 1 974, and October 1 973, respectively. These two fish had moved nearly 1 ,000 nautical miles ( 1 ,850 km). Two mer off northeastern North America Since 1961, 583 giant bluefin tuna have been tagged and released from traps in St. Margaret's Bay near Halifax, Nova Scotia, (Beckett 1970, Caddy and Burnett 1975, Burnett 1977), and 22 of these tags have been returned (Table 29, Figure 78). Ten of the recaptures were local (in St. Margaret's Bay), eight in the release season and t\vo in the following sum- mer The other 12 recoveries indi- cated longer migrations (Table 30), one to the area off Wedgeport, Nova Scotia, four to New England waters and seven into the Gulf of St. Lawrence. The fish recaptured off Wedgeport had been at liberty for only 49 days and four of those recap- tured in the Gulf of St Lawrence for from 63 to 84 days, so that all five migrations were "direct". The four fish whose tags were recovered off New England had been at large for from 12 to 35 months and the other three which were recaptured in the 100° 95" 90- 8b" 80° 50" 50' 15* 35" - 30" - 25' - 20" - LEGEND O RELEASE LOCALITY • BECAPTUBE LOCALITY (") RtTUHN NUMBER RETURN RELEASE RECAPTURE NUMBER DATE SIZE » DATE SIZE » X '«« (IStk^) at 'TO 194I19 Sl'rOIZOAhg) sa'70ir9k^ E ri IIBI kal m. '7i VOfcq SI '73(2?Sh9l sn'r« 2S6k« z '71 1182 kf) m 'rs 2SO«9 SI'7M29«ke) IE '77 TCOcn WEIGHTS ARE FOR WHOLE FiSH FIGURES IN PARENrMFSES WEHE ESTIMAIFO Figure 76 Geographic distribution of bluefin tag relea.se and recapture data from Bahamas tagging studies. 116 Table 30 Release and recovery data for bluefin tuna tagged in St. Margaret's Bay, Nova Scotia and recovered elsewhere, with size at recapture and months at liberty Recapture Return Release Numl )er Location Date 1 St. Margaret's Bay VIII- 1 966 44° 37' N 64° 00' W 2 St. Margaret's Bay VII-1966 44° 37' N 64° 00' W 3 St. Margaret's Bay VII- 1967 44° 37' N 64° 00' W 4 St Margaret's Bay VII- 1967 44° 37' N 64° 00' W 5 St. Margaret's Bay VIII- 1969 44° 37' N 64° 02' W 6 St. Margaret's Bay IX -1971 44° 37' N 64° 03' W 7 St. Margaret's Bay VII- 1971 44° 39' N 63° 59' W 8 St. Margaret's Bay VI- 1975 44° 37' N 64° 02' W 9 St. Margaret's Bay VI-1975 44° 30' N 63° 56' W 10 St. Margaret's Bay VI-1975 44° 36' N 64° 03' W 11 St. Margaret's Bay VII- 1971 44° 38' N 63° 59' W 12 St Margaret's Bay VI- 1976 44° 38' N 64° 00' W Location Date Weight Liberty Cape Cod, Massachusetts VIII- 1967 Cape Cod, Massachusetts VIII- 1967 Wedgeport, Nova Scotia IX- 1967 Cape Cod, Ma.ssachusetLs VI- 1969 Prince Edward Island, Canada VII- 1974 46° 40' N 63° 50' W Cape Elizabeth, Maine VIII-1974 43 30' N 70° 06' W Chaleur Bay, New Brunswick VIII- 1975 48° 08' N 64° 56' W Prince Edward Island, Canada IX-1975 47° 06' N 63° 55' W Prince Edward Island, Canada IX-1975 46° 29' N 62° or W Prince Edward Island, Canada IX-1975 46° 31' N 62° 07' W Prince Edward Island, Canada X-1976 46° 28' N 62° 44' W Chaleur Bay, New Brunswick VIII- 1 976 48° 09' N 64° 54' W 11.3 — 13.5 — 1.6 — 23.3 278 60.0 — 34.7 412 49.8 388 2.4 301 2.7 374 28 427 62.8 395 2.1 others travelled from Notre Dame Bay to St. Margaret's Bay, Nova Scotia, one in 47 months and the other in 59 months One tuna moved from Notre Dame Bay m August 1 972 to the southern extremity of the Grand Banks in December 1973. The times at liberty for the recaptured fish var- ied from a few days to nearly six years. Cooperating sport fishermen also marked 1 1 7 giant tunas in New En- gland waters, and 17 returns have resulted (Table 32) All but one re- capture were local, after penods of from a few days to about three years. The exception was a fish released July 22, 1974, in Massachusetts Bay off Boston and recaptured May 13, 1 975, in the Gulf of Mexico off south- western Florida (Figure 79) All the giant bluefin marked in New England and Canadian waters were tagged during the summer-fall feeding sea- son (July-October), and all but two of the recaptures occurred in the same general area and season This group of recaptures shows that large blue- fin move freely within the waters be- tween Cape Cod and Newfoundland, and return to them year af\er year. Evidently fish off northern New England and off Canada arc of a single stock, or of stocks which are only partially separated The only "direct" migrations, however, have been one from St. Margaret's Bay to Wedgeport and four from that Bay mto the Gulf of St Laurence. None of the recorded movements from Nova Scotia to New England and from Newfoundland to Nova Scotia or New England have occurred within a single season The two returns from outsule Ca- nadian and New England coastal wa- ters provide more significant indica- tions of migratory patterns The De- cember recapture, at the southern end of the Grand Banks, of a giant blue- fin which had been tagged in New- foundland coastal waters 16 months previously may show the initiation of a southerly migratory route even though it was not a direct migration. The migration from New England coastal waters in July to the Gulf of Mexico in the following May pro- vides the first definite and "direct" connection between the giant bluefin 117 Table 31 Releases and returns for giant (over 120 kg) blucfin tuna tagged in Newfoundland waters, by year of release, months at large, and area* of recaptuic Year 1 Releases Returns by Months at Large Total 0-5.9 6-17.9 18-29.9 30-41.9 42-53.9 >54 1962 6 1963 3 1964 41 1965 49 1966 49 1967 6 1968 217 IL IL 2L 1969 195 IN IN 1970 96 11, 1M,1N 3i.MN 1971 94 IM IM 1972 112 IG IG 1973 4 1974 14 1975 1976 Unknown 1 IL ♦Areas . L = Local, M = Massachusetts, N = Nova Scotia, G = Grand Banks which feed in New England and Ca- nadian waters in summer and early fall and those which spawn in the Gulf of Mexico in spring (Section VD3) As noted previously in this section, an "indirect" northward mi- gration from the Straits of Flonda to Massachusetts coastal waters has also been recorded. Another important tag return, which was not from the above re- lease groups, connects the feeding area olf New England and Canada with the spawning area for giants in the Gulf of Mexico. This fish, which was tagged when it was in the small size group, was released August 4, 1 966, 38 nautical miles (70 km) south by west of Martha's Vineyard, Mas- sachusetts, and was recaptured April 18, 1976, in the north central Gulf of Mexico, 135 nautical miles (250 km) south by east of the Mississippi River entrance (Figure 79). Its length when released was about 56 cm and af\er nearly 10 years at liberty, the longest such period of which we have knowl- edge for a tagged Atlantic bluefin tuna, it was reported as 2 1 8 cm. Thus the fish was age 1 + when tagged, and age 1 1 , or slightly less when recap- tured Although this fish was in the small size group when tagged, it must have been a giant when it made its last southward migration. Since less than 1 percent of the returns from bluefin tagged in New England wa- ters have been from outside the west- em North Atlantic, it is highly prob- able that this fmal migration was from the New England or Canadian feed- ing grounds The main significance of this return is the establishment of a connection between the nursery area south of Cape Cod and the spawning grounds in the south Moreover, it most probably also represents the sea- sonal migration of large bluefin from the northwestern Atlantic feeding area to their spawning areas This IS the third important link between the spawning area in the Gulf of Mexico and the summer feeding grounds off the noitheastem United States As noted previously m this section, a large bluefin made a "di- rect" migration from waters north of Cape Cod into the Gulf of Mexico, where it was recaptured during the spawning season. Another fish made an "indirect" migration, which might well have originated in the Gulf of Mexico, from the Straits of Florida, where po.st-spawning bluefin are prevalent, to New England waters. Thus tag returns have produced considerable information on move- ments of large bluefin within the western North Atlantic, and from it 50' Yin '72- -xn' 73 40* - 30- - LEGEND RELEASE LOCALITY RECAPTURE LOCALITY - LESS THAN '0 MONTHS AT LIBERTY — MORE THAN^OMONTMS AT LIBERTY RELEASE DATE RECAPTURE DATE II'64 • *<7rf LiNis joiNine DCLCAses AMOneruPfis ARC DIACKiyU'TIC AMD ARC MOT INTEMOCD TO RCRRCSCMT kneiATIOM ROOTCS 80* SO' 50* 40* Figure 77. Geographic disU-ibution of bluefin tag relea.se and recapture data from Newfoundland tagging studies 118 78 TO 6S 60 53 50 45 i r 80 r-^— J f/ "-? :»^ ""^ Ail^^*S — *?i\ JT * ^©'^mJwi ^^ *V/ 1 (' J 1 fe.wi - \ GRAND J BANKS { 4S BOSTON Ojg A / LEGEND 40 1 55 ^4'' -y MUA9C L0C4UTT • RECAPTURE LOC'LITy LESS TM»H 10 W0NTH8 «T LIBIBTT — — WORC THAN 10 HOMTNS AT LieCRTT f\ CiRCLtO WOMKBS RCPREUNT INOIVIOIML *-* ri»M «fCR"(0 ID IK T«Le Am oiAStumiuTic amo ami mot inreitoto TO TtCPHeSEHT MIC»£r/Om HCt/TTS Figure 78. Geographic distribution of bluefin tag release and recapture data from Nova Scotian tagging studies. into the eastern North Atlantic and the South Atlantic. Although con- nections have been established be- tween spring spawning areas in the Gulf of Mexico and the Straits of Florida and summer nursery and feed- ing areas off northwestern North Amenca, a complete migratory pat- tern has not emerged. We shall use the temporal and areal distribution of Table 32 Releases and returns for giant (>122 kg) bluefin tuna, Thunnus thynnus, tagged in New England coastal waters by years of release and time at large. Year 1 Releases Returns by Month at Large Total % 0-5.9 6 -17.9 18 - 29.9 30 -41.9 1966 2 1967 1968 6 1 1 167 1969 1 1970 4 1971 10 1 1 10 1972 17 1 1 2 4 235 1973 15a 5 1 1 7 46.6 1974 9" 2 2 4 44.4 1975 19 1976 34 ■Includes 6 releases of, and 4 1973 and 1 1975 returns from fish tagged while free swimming. ""Includes 3 releases of, and 3 returns from, fish tagged while free swimming longline catches and the observed re- lationships of the bluefm's distribu- tion and migratory habits to the ocean currents in completmg a hypotheti- cal model for the migrations of large western Atlantic bluefin. The limits of the western Atlan- tic area in which large bluefin spawn have not been established, but the spawning grounds may mclude most of the deep (over 200 m) waters be- tween latitudes 18°N and 33°N, and longitude 17°W and the coast (see Section VD3) The only thoroughly documented migration of large bluefin within this supposed spawning area is their northward passage through the Straits of Florida in May and June. This movement may be traced visually from boats or aircraft for about 65 nautical miles (120 km) along the northwestern edge of the Great Bahama Bank from Orange Cay to Great Isaac (Figure 29) when condi- tions are favorable (Farrington 1939, personal observations from U.S. Coast Guard Aircraft by L. R. Rivas, H. A. Schuck and F. J. Mather in May-June 1951, Rivas 1955, Mather 1964b). A continuation of this mi- gration along the 40 nautical mile (74 km) western edge of the Little Bahama Bank has also been observed (personal observations by the semor author at Memory Rock and Matanilla Shoal, June 1966, catches off West End, Grand Bahama, per- sonal communications) The senior author also sighted and baited schools of giant bluefin which were traveling northward on the surface at the west- ern end of the Northwest Providence Channel, between the Great and Little Bahama Banks, in May-June 1968. This unusual observation was a good indication of the continuity of this migratory route. Other migrations within the spawning areas may be deduced from the times and locations of catches, sightings and/or the destruction of equipment of anglers fishing for billfishes and other less powerful game, as well as occasional appear- ances of schools of bluefin. The Windward Pa.ssage appears to be a focal point in the migrations of giant bluefin into the spawning aiea. Large fish, including some with 119 50' 45' 40" 35' 30' 25* 20" 15» 95° — T— 85" 1^ LEGEND o RELEASE LOCALITY • RECAPTURE LOCALITY O RETURN NUMBER ) RETUHN REL EASE RECAPTURE NUMBER oaTT Sl^t • DATE SiZe» 1 sn'ee |7A-81eml II'75 200>g 2 sn'M IZSJkfl 17! 250kg,250OT 3 SS'66 (53-59c«. ni'76 Jiem 4 SI'TS IE'77 263tm 5 SIIl'76 I'77 e SI'75 m*7B 280CI. 7 im'77 BETS 340lig. 262cm « 2n'77 Z'78 365lig.271c