iss-.!*:**** 3 aR a ui h U.S. DEPARTMENT OF COMMERCE Philip M. Klutznick, Secretary National Oceanic and Atmospheric Administration Richard A. Frank, Administrator National Marine Fisheries Service Terry L. Leitzell, Assistant Administrator for Fisheries The National Marine Fisheries Service (NMFS) does not approve, rec- ommend or endorse any proprietary product or proprietary material mentioned in this publication. No reference shall be made to NMFS, or to this publication furnished by NMFS, in any advertising or sales pro- motion 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 an intent to cause directly or indirectly the advertised product to be used or purchased because of this NMFS publication. CONTENTS Introduction 1 1 Identity 1 1.1 Nomenclature 1 •1.11 Valid 1.12 Synonymy 1 1.2 Taxonomy 1 1.21 Affinities 1 1.22 Taxonomic status 4 1.23 Subspecies 4 1.24 Standard common names, vernacular names 4 1.3 Morphology 5 1.31 External morphology 5 *1.32 Cytomorphology 1.33 Protein specificity 7 2 Distribution 12 2.1 Total area 12 2.2 Differential distribution 13 2.21 Spawn, larvae, and juveniles 13 2.22 Adults 14 2.3 Determinants of distribution changes 14 2.4 Hybridization 15 2.41 Hybrids; frequency of hybridization, species with which hybridization occurs; methods of hybridization 15 *2.42 Influence of natural hybridization in ecology and morphology 3 Bionomics and life history 15 3.1 Reproduction 15 3.11 Sexuality 15 3.12 Maturity 15 3.13 Mating 16 3.14 Fertilization 16 3.15 Gonads 16 3.16 Spawning 16 3.17 Spawn 22 3.2 Preadult phase 26 3.21 Embryonic phase 26 3.22 Larvae phase 26 3.23 Adolescent phase 27 3.3 Adult phase 31 *3.31 Longevity 3.32 Hardiness 31 3.33 Competitors 31 3.34 Predators 32 3.35 Parasites, diseases, injuries, and abnormalities 33 3.4 Nutrition and growth 34 3.41 Feeding 34 3.42 Food 34 3.43 Growth rate 36 3.44 Metabolism 37 3.5 Behavior 38 3.51 Migrations and local movements 38 3.52 Schooling 38 3.53 Responses to stimuli 39 4 Population 41 4.1 Structure 41 4.11 Sex ratio 41 *4.12 Age composition 4.13 Size composition 43 iii 4.2 Abundance and density 45 4.21 Average abundance 45 *4.22 Changes in abundance *4.23 Average density *4.24 Changes in density 4.3 Natality and recruitment 48 4.31 Reproduction rates 48 *4.32 Factors affecting reproduction 4.33 Recruitment 48 4.4 Mortality and morbidity * 48 *4.41 Mortality rates 4.42 Factors causing or affecting mortality 48 *4.43 Factors affecting morbidity *4.44 Relation of morbidity to mortality rates *4.5 Dynamics of population *4.6 The population in the community and the ecosystem 5 Exploitation 48 5.1 Fishing equipment 48 5.11 Gears 48 5.12 Boats 50 5.2 Fishing areas 51 5.21 General geographic distribution 51 5.22 Geographic ranges 51 5.23 Depth ranges 51 *5.24 Conditions of the grounds 5.3 Fishing seasons 52 5.31 General pattern of seasons 52 5.32 Dates of beginning, peak, and end of season ~ 52 5.33 Variation in date or duration of season 52 5.4 Fishing operation and results 53 5.41 Effort and intensity 53 5.42 Selectivity 53 5.43 Catches 54 Literature cited 57 *No information available. Synopsis of Biological Data On Frigate Tuna, Auxis thazard, and Bullet Tuna, A. rochei Richard N. Uchida ABSTRACT This synopsis of biological and technical data on frigate tuna, Auxis thazard, and bullet tuna, A. rochei, includes information on identity, distribution, bionomics, life history, population, and ex- ploitation. Over 200 published and unpublished reports, up to and including those published in 1978, are covered. INTRODUCTION Unlike a number of tuna species in the world's oceans which are heavily exploited and possibly are being harvested near the upper limit of rational utilization, frigate mackerels (or frigate and bullet tunas; see Klawe 1977), the most primitive genus of the higher tunas (tribe Thunnini), give every indication of being an underutilized fishery resource. In planning for the ra- tional utilization of this resource, it is imperative to have on hand information concerning the biology of the species and estimates of present and potential catches. The purposes of this paper, therefore, are to update our current knowledge of the species by bringing together and abstracting all the available information on the biology and fisheries for frigate and bullet tunas and to review and evaluate in depth the results of past research, indicating in particular where conflicting evidence exists. Scomber thazard Lacepede 1802 (original description; off coast of New Guinea) Scomber bisus Rafinesque 1810 (original description; Palermo) Thynnus rocheanus Risso 1826 (original description; Nice) Auxis taso Cuvier and Valenciennes 1831 (original de- scription; New Guinea) Auxis vulgaris Cuvier and Valenciennes 1831 (original description; Mediterranean) Auxis tapeinosoma Bleeker 1854 (original description; Japan) Auxis thynnoides Bleeker 1855 (original description; Ter- nate) Auxis rochei. Gunther 1860 Auxis thazard. Jordan and Evermann 1896 Auxis hira Kishinouye 1915 (original description; Japan) Auxis maru Kishinouye 1915 (original description; Japan) Auxis bisus. Cadenat 1950 1 IDENTITY 1.1 Nomenclature 1.2 Taxonomy 1.21 Affinities Auxis thazard (Lacepede) 1802 Auxis rochei (Risso) 1810 1.12 Synonymy It it not possible at this time to assign all the various names that have appeared in the literature to either thazard or rochei. The following names are from Rosa (1950), de Beaufort and Chapman (1951), Collingnon (I960), 2 Williams (1963), 3 Idyll and de Sylva (1963), and Jones (1963) and are listed in chronological order. 'Southwest Fisheries Center, National Marine Fisheries Service, XOAA, P.O. Box 3830, Honolulu. HI 96812. "Collingnon, J. 1960. Report on Auxis thazard in the eastern Atlan- tic. [InFr.] CCTA Symposium on Thunnidae, Dakar, 12-27 December 1960, 5 p. (Mimeogr.) (Engl, transl. available Southwest Fish. Cent., Natl. Mar. Fish. Serv., NOAA, Honolulu, HI 96812.) Williams, F. 1960. A symposium of existing knowledge on the lishes on the genus Auxis Cuvier, 1829 in the Indian Ocean. CCTA Kingdom Animalia Phylum Chordata Subphylum Vertebrata Superclass Gnathostomata Class Osteichthyes Subclass Actinopterygii Division Teleostei Cohort Acanthopterygii Order Perciformes Suborder Scombroidei Family Scombridae Subfamily Scombrinae Tribe Tunnini (Thunnus, Katsuwonus, Euthynnus, Auxis) Symposium on Thunnidae, Dakar, 12-27 December 1960. 13 p. (Mimeogr.) Genus Auxis Cuvier 1829 (type-species: Scomber rochei Risso) by subsequent selection by Gill (1862). The description of the genus Auxis under "Les Scombres" first appeared in Cuvier's Regne Animal in 1829. From the time of Cuvier's work to the present, however, the genus Auxis has had several reclassifica- tions. For example, in classifying the genus Auxis, Kishi- nouye (1915) first included it in the family Thunnidae. However, in a later work (1917), he placed Thunnidae and Katsuwonidae in a new order called Plecostei and placed the families Scombridae and Cybiidae in the order Teleostei. The genus Thunnus fell under Thun- nidae and the genera Katsuwonus, Euthynnus, and Auxis came under Katsuwonidae (Kishinouye 1923). The primary characteristic subcutaneous blood vessels; a secondary characteristic of Plecostei was the presence of well-developed subcutaneous blood vessels; a secondary characteristic was the development of dark red lateral tissues in relation to the subcutaneous blood vessels. Many scientists disagreed with Kishinouye's new order and its subdivision. Takahashi (1924) argued that Plecostei was established only on partial differences in the highly variable vascular system and cannot exist on an equal status with the other four orders of Teleo- stomi. Jordan (192.3) and Herre (1953) placed the scombroid fishes in two families — Scombridae and Thunnidae. Fraser-Brunner (1950), on the other hand, rejected any division of the family Scombridae arguing that attempts to subdivide this family "have resulted in arrangements which are artificial and have left the clas- sification in an uneasy, shifting state." de Sylva (1955), Collette and Gibbs (1963b), and most other re- cent workers recognize a single family Scombridae with various subdivisions. Collette and Chao (1975) placed Auxis in the tribe Thunnini of the subfamily Scombri- nae. The following description of the genus Auxis is from Jordan and Evermann (1905): "Body oblong, plump, most naked posteriorly, anteriorly covered with small scales, those of the pectoral region enlarged, forming a corselet; snout very short, conical, scarcely compressed; mouth rather small, the jaws equal; teeth very small, mostly in a single series, on the jaws only; tail very slender, depressed, with a rather large keel on each side; first dorsal short, separated from the second by a consi- derable interspace; second dorsal and anal small, each with 7 or 8 finlets; pectorals and ventrals small; no air- bladder; branchiostegals 7; pyloric coeca dentritical; gill- rakers very long and slender, numerous; vertebrae 39 in number, peculiarly modified . . . ." Auxis is the most primitive genus among the higher tunas that have developed a prootic pit and a partial subcutaneous circulatory system (Collette and Gibbs 1963b). All the Thunnini, except Auxis, have a common cutaneous artery that divides into dorsal and ventral branches lateral to the aorta; in Auxis, however, the dor- sal and ventral branches originate separately with the latter being very poorly developed (Collette 1978). The haemal spines of the thoracic vertebrae do not form a haemal arch and the first vertebra is not sutured to the cranium as in the higher members of Thunnini. Also, compared with Euthynnus, Katsuwonus, and Thunnus, Auxis lacks the frontoparietal fenestra, which is an ad- ditional pair of openings present in the cranium and has a lateral countercurrent system for heat exchange that is not as well developed phylogenetically (Collette 1978). All members of the tribe Thunnini have a swim bladder as juveniles; however, the bladder degenerate with growth in Auxis, Euthynnus, and Katsuwonus. Several other characters distinguish members of the genus Auxis, the smallest of the higher tunas, from other scombrids (Collette and Gibbs 1963b; Fitch and Roedel 1963; Williams 1963). In Auxis, there is a single interpel- vic process which is between and about as long as the pel- vic fins. In other tunas, but not in other scombrids, the interpelvic process is bifurcate and much less than half the length of the pelvic fins (it is single but small in Grammatorcynus and Gymnosarda). The number of dor- sal and anal finlets (seven or eight) distinguishes Auxis from Scomber, which has only five of each. Auxis can also be distinguished from Rastrelliger by the lack of a corselet and a body entirely covered with moderate-sized scales in the latter. Species — Auxis thazard (Lacepede) 1802 The following description of A. thazard (Fig. 1, upper photo) is from Williams (1963). Major morphological features described under the genus are omitted from the account of the species. "Depth 3.9 to 4.5, head 3.2 to 3.8 in standard length. Eye 5.0 to 5.85 in head, 1.25 to 1.66 in snout and 1.25 to 1.7 in the flatly rounded interorbital space. Snout 3.62 to 4 in head. Maxilla reaches to a point under the anterior half of the eye and is 3 in head. Single row of small pointed teeth in each jaw, none on palate. Jaws almost equal. First and second dorsal spines subequal, equal to snout and eye; following spines rapidly decreasing in size, eighth usually shorter than the pupil. Second dorsal fin very low, about three times its base distant from first dorsal; first ray of second dorsal about 5 in head. Anal similar to second dorsal, first ray about 5.2 in head. Pec- torals short, roughly triangular, about 2 in head and shorter than postorbital; origin of pectoral before that of first dorsal. Pelvics thoracic, about 2.5 in head, origin somewhat behind that of pectorals. Caudal lunate, upper lobe about 1.8 in head. Body naked except for the corse- let of scales anteriorly. Rear margin of the corselet runs from base of second dorsal to above end of pectoral; thence there is a posterior prolongation of the corselet along the lateral line; below the pectoral tips the corse- let margin curves to above the pelvic base from where it turns posteriorly and finishes well behind the tips of the pelvics. The prolongation of the corselet along the lateral line tapers abruptly between first and second dorsal fins, and under the origin of the second dorsal is not more than 4 irregular scale rows wide. Scales large and imbricated above pectoral base. Gill rakers about 1.75 in length of gill filaments." Figure 1. — Auxis thazard (upper photo) from the eastern Pacific, collected at Morgan Bank, Baja California, and A. rochei (lower photo) from near Santa Catalina Island, Calif. (Fitch and Roedel 1963) Species - Auxis rochei (Risso) 1810 The description of A. rochei (Fig. 1, lower photo) is from Jones (1958), who originally referred to the specimen as A. tapeinosoma. "Body robust, rounded, almost circular in cross- section. Dorsal outline moderately and evenly curved. Ventral outline evenly curved when fresh but slightly flattened abdominally after preservation in formalin. "Height 5.38 in standard and 5.55 in furcal [= fork] length. Head 3.77 in standard and 3.88 in furcal length. Snout, pointed 3.9 in the head, longer than eye diam- eter. Eye 4.91 in the head, 1.27 in the snout, 1.27 in the almost flattened interorbital space. Mouth moderate, oblique, end of maxillary reaching vertical from anterior margin of eye. Jaws nearly equal, the lower jaw pro- jecting almost imperceptibly beyond the upper. Teeth small, pointed in a single row on both jaws, none on palate. Branchiostegals 7. Gill rakers long and slender, 45 in first gill arch. "Two dorsal fins separated by interspace slightly shorter than head length. First dorsal roughly triangular with 10 spines, anterior spine longest. Second dorsal small with 13 rays. Dorsal finlets 8. Anal fin small, with 2 spines, 11 rays. Anal finlets 7. Pectorals roughly tri- angular, reaching vertical from the base of first ray of first dorsal. Ventral thoracic, axillary scales equal in length to ventrals. "Body naked except for the corselet of scales which taper gradually to 9-10 irregular scale rows at vertical through second dorsal and end as a narrow line at ver- tical below second dorsal finlet. Scales large and imbri- cated above pectoral base. Caudal peduncle slender with feebly developed lateral keels. Lateral line somewhat un- dulating and without a distinct arch." 1.22 Taxonomic status A specimen of frigate tuna, collected by Commerson off New Guinea in 1768, was first described by Lacepede in 1802 as Scomber thazard. In 1810, Risso and Rafines- que, working independently, named the Mediterranean form Scomber rochei and Scomber bisus, respectively. Gill (1862) subsequently designated rochei as the type- species for Auxis Cuvier. Systematists generally followed Gunther (1860) and Jordan and Gilbert (1882) in using Risso's rochei instead of Rafinesque's bisus, but no one has indicated which name was described first. Further- more, the International Commission of Zoological Nomenclature has never ruled as to which name, rochei or bisus, is valid (Fitch and Roedel 1963). Over the years, other names appeared in the literature. These include Thynnus rocheanus Risso, 1826 (Mediter- ranean), A. vulgaris Cuvier and Valenciennes, 1831 (Mediterranean), A. taso Cuvier and Valenciennes, 1831 (New Guinea), A. tapeinosoma Bleeker, 1854 (Japan), and A. thynnoides Bleeker, 1855 (Ternate). In 1915, Kishinouye described two additional names. The species he named hira had a short corselet which ended slightly posterior to the pectoral fin. The other he named maru had a long corselet which extended to the anal fin. Kishi- nouye stated that the maru is probably the same species as thazard, but could not say whether thazard corre- sponded to hira or maru. Furthermore, he believed that Bleeker's (1854) tapeinosoma could be maru, but the figure and description of this species were unclear and he was unable to make a positive identification. Therefore, Kishinouye described them as new species, hira and maru. Fraser-Brunner (1950) disagreed with Kishinouye and recognized only a single, worldwide species, thazard, but others, such as Wade (1949), Cadenat (1950), and Jones (1958), recognized two species of Auxis. Wade (1949) used Bleeker's name tapeinosoma for a long-corseletted pacific form. Matsumoto (1959, 1960a), on the other hand, finding the nomenclature of the long-corseletted form confused, resurrected the name thynnoides. He argued that tapeinosoma of Wade (1949) and Herre and Herald (1951) appeared to be a misnomer because it is actually a short-corseletted form. Furthermore, Matsu- moto regarded hira and maru of Kishinouye and tape- inosoma of Bleeker as synonyms of the short-corseletted, worldwide thazard. 1.23 Subspecies None. 1.24 The standard common names, vernacular names The standard common names and vernacular names of A. thazard were abstracted from the lists published in Fiedler (1945), Rosa (1950), Collingnon (1960— see foot- note 2), FAO (1960, 1976), Idyll and de Sylva (1963), Uchida (1963), Williams (1963), Institute del mar del Peru (1971 4 ), Miyake and Hayasi (1972), Klawe (1977), and Roberts et al. (1977). Country Common and vernacular names Algeria Auxide, Scunno, Bisu, Melva, Mel- vara Angola Jedeu Australia Frigate mackerel, Leadenall Brazil Bonito cachorro British Guiana Frigate mackerel, Blowgoat British West Frigate mackerel, Blowgoat, Round- Indies belly bonito Canada Frigate mackerel, Thazard Denmark Auxide East Africa Frigate mackerel, Sehewa (Kiswa- hili) also refers to Euthynnus and Katsuwonus sp. Ecuador Botellita France Auxide, Bonitou, Bounitou, Bouni- cou, Palamida, Auxide bise, Tazard, Bizet French Morocco Melva French West Melva Africa Ghana Okpopu, Odaabi, Poku-poku Gold Coast Frigate mackerel Greece Kopani Haiti Maquereau Hawaiian Is- Frigate mackerel, Keokeo, Mexican lands skipjack India Frigate mackerel, Churai, Urulan- churai, Kutteli-churai (Tamil) Israel Tuna nanasit Italy Strumbo, Tambarello, Strombo, Ivory Coast Japan Auxis sp. A. thazard A. hira Korea Madeira Malta Island Mexico Scurmo, Tombarello, Tamburella, Tambarella, Sgionfetta, Tumbarel, Bisu, M'pisu, Pisantuni, Mazzita, Sangulu, Culariau, Sgamiru, Tun- nacchiu, Biso Boku-boku, Poku-poku, Bongu Sodagatsuwo Hirasoda, Hirasodakatusuo, Hirame- jika, Obosogatsuwo, Shibuwa, Soma, Suma, Oboso (Hirasohda, Hiraga- tsuo, and Hiramedika are varia- tions in spelling of some of the above names) Mul-chi-da-rae, Mul-chi, Mu-tae-da- raeng, Mog-man-dung-i Chapouto Tombrell, Mazzita, Tombitombi, Zgamirru Bonito 'Institute del Mar del Peru. 1971. Report of the "Institute del mer del Peru" and other institutes of fishery investigations from South America for the fourth session of the FAO panel of experts for the facili- tation of tuna research. La Jolla, Calif., 8-13 November 1971, FIR:EPFTR/71/Inf. 12, 26 p. (Mimeogr.) Melva Frigate tuna Auxis Deho Melva, Macarela Frigate mackerel, Tunungan, Man- ko, Mangko (Marinao, Samal, Vis- aya, and Tao Sug) Gayado, Judea, Serra, Cachorra, Bo- nito Tongkol Bonite folle Tubani (Somali) Frigate mackerel Melva, Visol, Melvara, Bis, Bonito del Norte, Macaela Melva Frigate mackerel, Rogodwa (Sinhalese), Alagoduwa Auxide Chien yu PlaO Gobene Frigate mackerel, Boo hoo Auksida, Makrelevyi tunets Plain bonito Frigate mackerel, Bullet mackerel, Boo hoo, Frigate tuna Cabana negra Trupac, Tunjcic, Rumbac Morocco New Zealand Norway Papua New Guinea Peru Phillipine Is- lands Portugal Sarawak Seychelles Somalia (Mi- jurtein coast) South Africa Spain Spanish Morocco Sri Lanka (Ceylon) Sweden Taiwan Thailand Turkey Union of South Africa Union of Soviet Socialist Re- public United King- dom Unites States Venezuela Yugoslavia International organization Food and Agri- Frigate tuna, Auxides, Melvas culture Organi- zation of the United Nations (FAO) International Frigate tuna, Auxide, Melva Commission for the Conserva- tion of Atlantic Tunas (ICCAT) Jones (1963), Uchida (1963), Institute del Mar del Peru (see footnote 4), FAO (1976), and Klawe (1977) pro- vided the following common and vernacular names of A. rochei = A. tapeinosoma = A. thynnoides = A. maru. Country Common and vernacular names Australia Long corseletted frigate mackerel, Maru frigate mackerel Ecuador India: Malayalam (North) Malayalam (South) Tamil Japan Peru Union of Soviet Socialist Re- public United States International organization FAO ICCAT Botellita Kuttichoora (means small tuna and generally applied to A. thazard and young of E. af finis) Urulan-choora (means rounded tuna and applied to A. thazard also from which this species is not generally distinguished) Eli-choorai (means ratlike tuna), Kutteli-choorai (means small rat- like tuna) Marusoda, Marusodakatsuo. Maruga- tsuwo, Marumejika, Magatsuwo, Manba, Mandara, Chiboh, Dain- anpo, Nodoguro, Rohsoku, Subota, Uzawa. Maiika, Soku, Soda, Subo. (Marumedika, Marugatsuo, Maga- tsuo, and Dainanbo are variations in spelling of some of the above names.) Other names mentioned by Rosa (1950) are Kogatsuo, Kukarai, Kobukura Melva, Fragata Skumbrievyi tunets, Auksida Bullet mackerel, Bullet tuna Bullet tuna Bullet tuna To avoid confusion throughout the remainder of this paper, A. rochei will be used instead of A. tapeinosoma, A. thynnoides, or A. maru to describe the long-corse- letted form. For the short-corseletted form, A. thazard will be used, but there is considerable confusion in the literature because several authors used thazard in the belief that there was only a single worldwide species of Auxis (Fraser-Brunner 1950; Rivas 1951). This usage must be translated into Auxis spp. where reference is made to the Pacific forms and into A. rochei in the Atlan- tic based on Fitch and Roedel's (1963) interpretation or Auxis spp. based on Richards and Randall's (1967) docu- mentation of A. thazard in the Atlantic. 1.3 Morphology 1.31 External morphology The fin ray counts, together with the gill raker and vertebral counts, are given for A. thazard and A. rochei in Tables 1 and 2, respectively. It should be noted that some investigators did not separate the two species of Auxis at the time the meristic counts were made and, therefore, their counts have not been included. One of the sources of confusion in identifying Auxis to Table 1. — Meristic characters of the short-corseletted frigate tuna, Auxis thazard, by various investigators. First Second Dorsal Anal Anal Verte- Branchi- dorsal fin dorsal fin finlets fin finlets Gill rakers brae ostegals Hawaii: Matsumoto (1960a) XI 10-12 8 13 7 (9-10) + 1 + (28-31) = 39-42 20+19 — Yoshida and Nakamura (1965) — — — — — (9-10) + 1 + (29-31) = 39-42 — Japan: Kishinouye(1923) XXI 12 8 13 7 9+30 20+19 — Philippines: Wade (1949) X-XII 10-12 8 II 8-11 7 (9-10) + 1 + (27-32) = 37-43 7 India: Jones (1958) X 13 8 i i, 11 7 40 — Table 2. — Meristic characters of the long-corseletted bullet tuna, Auxis rochei, by various investigators >. First Second Dorsal Anal Anal Verte- dorsal fin dorsal fin finlets fin finlets Gill rakers brae Hawaii: Matsumoto (1960a) X-XI 10-11 8 12-13 7 (10-11) + 1 + (32-36) = 43-48 20+19 Yoshida and Nakamura (1965) — — — — — (10-11) + 1 + (33-37) = 44-49 — Japan: Kishinouye(1923) IX-X 10-12 8 13 7 10+36 20+19 Philippines: Wade (1949) X-XI 10-12 7-8 D , 10-12 7 (10-12) + 1 + (31-35) = 44-48 — Indonesia: De Beaufort and Chapman (1951) X 11 6-9 14 6-8 — — South Africa: Talbot (1964) x-xn 11 8 14 6-7 (9-10) + (32-34) — India: Jones (1963) XXI 13 8 13 7 (8-12) + (31-36) — species stems from an overlap in the width of the corse- let. Klawe (1963), lacking evidence to substantiate the presence of two distinct species of Auxis in the eastern Pacific, noted that there were intermediate forms of the short- and long-corseletted forms. Provisionally, he used A. thazard for frigate tuna from this region, but empha- sized the possibility that as samples accumulate, there may be sufficient evidence to substantiate the presence of two separate species. From the Indian Ocean, a few large adults of A. thazard from the southwest coast of In- dia were recognized as having a corselet that narrowed gradually somewhat as in A. rochei instead of one which tapered abruptly (Fig. 2) (Jones 1963). In an attempt to unravel the confusion involving this genus, Fitch and Roedel (1963) examined numerous adult Auxis, mostly from the Pacific, but failed to find any significant morphometric differences among A. thazard from the western, central, and eastern Pacific (Table 3). Based on gill raker counts, they concluded that the eastern Pacific population seems to be separ- able from those in the central and western Pacific (Table 4). For adult A. rochei, Fitch and Roedel found apparent differences in body measurements among areas (Table 5) and in the average number of scale rows in the corselet. In an earlier study, Matsumoto (1960a) observed that the number of scale rows in the corselet increased with fish length. But Fitch and Roedel showed that in addition to the positive relationship between these two variables, there was also an increase in the average number of scale Table 3.- -Selected measurements of Auxis thazard from three geo- graphical localities (Fitch and Roedel 1963). Western Central Eastern Pacific Pacific Pacific Number of specimens 10 14 60 Ranges in standard length (mm) 200-402 248-384 263-392 Ranges in percent of standard length: Head 26.5-28.9 27.5-29.4 27.4-29.5 Eve 4.7-5.5 4.8-5.5 4.2-5.1 Snout 6.3-7.5 6.4-7.4 5.7-6.8 Pectoral length 12.2-14.2 13.1-14.9 12.7-14.6 Snout to first dorsal 30.6-33.2 32.3-33.6 31.8-34.5 Snout to second dorsal 61.5-66.2 62.6-66.9 64.6-69.0 Snout to anal 67.8-72.2 68.3-72.1 68.3-74.2 Depth 21.5-25.2 20.6-25.3 22.4-26.4 Width 14.6-17.2 15.1-19.5 16.7-20.2 rows for similar-sized fish from west to east. Auxis rochei in the western Atlantic had the fewest scale rows whereas those in the eastern Pacific had the most. And for A. thazard, Tortonese (1965) added that the Mediterran- ean forms have characters that are not entirely identical to those of the Indo-Pacific forms and suggested that geo- graphical variations may be involved. As a result of their study, Fitch and Roedel (1963) ten- tatively recognized two valid species — A. thazard and A. rochei. A summary of external and internal charac- ters used by several investigators to differentiate the two species of Auxis is given in Table 6. ^ b/ (a) (b) (c) Figure 2. — Outlines of corselets drawn from actual specimens (Jones 1963). (a) Auxis rochei; (b) A. thazard - typical short- corseletted condition; (c) A. thazard - intermediate condition. Among internal characters that have been a source of confusion is the gill raker count. Wade (1949) and Herre and Herald (1951) pointed out that A. thazard in the western Pacific normally have fewer gill rakers than A. rochei. And A. thazard taken in the eastern Pacific and the Indian Ocean also have counts that are definitely lower (Mead 1951; Jones 1963). Godsil (1954), on the other hand, reported that all the A. thazard taken off Baja California and the Galapagos Islands have high gill raker counts similar to those reported for A. rochei by Wade (1949) and Herre and Herald (1951). Matsumoto (1959), however, believed that the number of gill rakers, by itself, is not a reliable character in identifying the two species of Auxis. But Jones and Silas (1964) suggested that the gill raker counts could be useful; in case of doubtful identification from external characters, a com- bination of gill raker counts and corselet width should facilitate specific identity. Jones and Silas (1964) have used body cross section as an aid to identification but Collette and Gibbs (1963a) have warned that this character is difficult to use (Fig. 3). Also suggested by Jones and Silas was the position of the visceral organs (Fig. 4). Godsil (1954) and Yoshida and Nakamura (1965) noted that other prominant dif- ferences between the two species occur in the skeletal structure. In A. thazard, the temporal crests diverge an- teriorly so that they are not parallel to one another (Fig. 5) and the width of the skull is wider in relation to body length (Fig. 6). Yoshida and Nakamura also noted that the length of the anterior branch of the haemal processes was longer and touching on the preceding arch of the 24th to the 28th vertebrae in A. thazard (Fig. 7). 1.33 Protein specificity Taniguchi and Nakamura (1970) examined muscle protein of A. thazard and A. rochei by the cellulose acetate electrophoretic method to determine whether specific divergence occurred between species. They found five components in the electropherograms of both species, but some components were not common to both. The genus Auxis, they concluded, contains two distinct species based on external and internal morphological characters although they are closely related. To analyze muscle protein polymorphism in Auxis col- lected from the coastal region of Kochi Prefecture, Japan, Taniguchi and Konishi (1971) used starch-gel electrophoresis and detected differences in protein specificity between A. thazard and A. rochei. They con- cluded that whereas no individual variation could be rec- ognized in electropherograms of 11 specimens of A. Table 4. -Comparison of gill raker counts for 84 Auxis thazard from three geographical localities (Fitch and Roedel 1963). 8 Upper limb 9 10 11 Nu 12 mber of rakers on Center first arch Lower limb Area 1 28 29 30 31 32 33 34 35 36 Western Pacific Central Pacific Eastern Pacific 1 7 6 4 1 8 31 1 24 1 10 1 14 — 60 — 1 3 5 2 1 — 6 5— — 1 5 11 14 19 9 1 Total number of rakers Area 38 39 40 41 42 43 44 45 46 47 Western Pacific Central Pacific Eastern Pacific 2 — 4 3 2 2 5 1 1 1 — 4 — — 5 8 13 15 11 7 'On all specimens the center raker has roots extending into both limbs. 7 Table 5.— Selected measurements of Auxis rochei from five geographic localities (Fitch and Roedel 1963). Western Eastern Western Central Eastern Atlantic Atlantic Pacific Pacific Pacific Number of specimens 24 3 6 2 28 Ranges in standard length (mm) 277-347 368-398 206-250 272-277 253-352 Ranges in percent of standard length: Head 26.3-28.0 26.1-26.9 26.0-27.8 27.4-27.8 26.4-28.1 Eye 4.5-5.1 4.6 4.8-5.4 5.1 4.5-5.3 Snout 6.0-7.3 6.3-6.6 5.8-6.8 6.6-7.2 a.0-6.6 Pectoral length 12.0-14.0 - 12.2-12.7 11.7-13.2 13.1-13.7 12.2-13.9 Snout to first dorsal 30.7-33.3 30.7-31.9 30.4-32.5 31.3-32.5 30.7-33.0 Snout to second dorsal 63.3-67.8 66.4-68.6 64.0-65.8 65.1-66.8 65.4-68.3 Snout to anal 68.2-73.9 71.1-72.4 68.8-72.8 71.3-71.5 68.0-74.0 Depth 20.5-24.2 23.2-23.6 19.8-21.4 20.6-21.0 22.1-24.3 Width 14.7-18.1 17.9-20.6 13.3-16.1 14.7-15.2 16.4-18.9 Ranges in scale rows 6-9 7-10 '9-13 '11-15 13-28 Matsumoto (1960a) gives 9-15 for Roedel's (1963) two central Pacific 9 western Pacific specimens and 16-18 for 20 from the central Pacific. Fitch and specimens were from Matsumoto's lot of 20. Table 6. — Characters used by several investigators to differentiate Auxis thazard from A. rochei (Godsil 1954; Fitch and Roedel 1963; Jones and Silas 1964; Yoshida and Nakamura 1965). Auxis thazard 1. Fifteen or more oblique to nearly horizontal dark wavy lines in bare area on each side of back. 2. Corselet of scales running along lateral line is, at most, three rows wide where it passes beneath second dorsal. 3. Pectoral fins extend beyond a vertical from anterior margin of patterned bare area on back. 4. Body compressed from side to side (Fig. 3). 5. Shape of abdominal cavity more oval. 6. Right lobe of liver makes a complete loop crossing over mid- ventral longitudinal axis (Fig. 4). 7. Stomach extends to slightly behind anal opening as does right lobe of liver. 8. Caecal mass occupies less space, spleen is smaller, and left lobe of liver relatively long. 9. Temporal crests diverge anteriorly and not parallel with each other (Fig. 5). 10. Skull wide relative to its length (Fig. 6). 11. Anterior branch of haemal processes long and touching preceding haemal arches on 24th to 28th vertebrae (Fig. 7). Auxis rochei Fifteen or more broad, nearly vertical dark bars on bare area on each side of back. Corselet with more than six rows of scales where it passes beneath second dorsal. Pectoral fins fail to reach vertical beneath the anterior end of the dorsal bare area. Body more rounded and robust. Shape of abdominal cavity dorsally compressed. Right lobe of liver shows no looping and hepatic vein not in line with midventral longitudinal axis. Stomach is short with distal end not reaching anal opening and right lobe of liver extends backward but does not sur- pass a line drawn from origin of anal fin. Caecal mass occupies more space, spleen is larger, and left lobe of liver relatively short. Temporal crests of skull parallel with each other and supra- occipital crest. Skull narrow relative to its length. Anterior branch of haemal processes short, fragile, and separated from preceding haemal arches. thazard, all 170 specimens of A. rochei fitted into one of three phenotypic protein patterns. They hypothesized that these three phenotypes are controlled by two codominant alleles. Furthermore, the distribution of the three phenotypes was independent of age and sex. Almost all proteinases found in animal meat exhibit pH optima in the acid range; however, Makinodan and Ikeda (1969), studying fish muscle protease, concluded that there are actually two types of proteinases in fish muscle — one acting in the acid pH range and the other in the slightly alkaline range. They found that the former occurred in all fish tested whereas the latter was only in fishes with white flesh except the cod, Gadus macro- cephalus. For red or slightly red-flesh fish such as alba- core, Thunnus alalunga; bullet tuna, A. rochei; common mackerel, Scomber japonicus; sardine, Sardinops melanosticta; yellowtail, Seriola quinqueradiata; and horse mackerel, Trachurus japonicus, proteinase activ- ity was either low or not present. In attempts to find an easier and faster method of identifying larval and postlarval tunas, Matsumoto (1960b) experimented with paper chromatography, an important technique used to identify chemical com- pounds. On the assumption that the free amino acids in the muscle tissues of fishes are hereditary, Matsumoto attempted to separate adult A. thazard from A. rochei but encountered difficulty in distinguishing them; how- ever, he was able to separate the two species from other tunas. >• Sharp and Pirages (1978) inferred the phylogeny of several species in the four tribes of the subfamily Scom- brinae, based on comparison of electrophoretic mobilities of several proteins. Calculating the percent- age of protein bands that are shared and that show similarity between species pairs, they concluded that Euthynnus lineatus is more primitive than A. thazard. Placement of A. thazard above E. lineatus in the phylogeny is supported by the higher affinity of Auxis to both Thunnus albacares and E. lineatus than E. lineatus exhibits to any of the other Thunnini. 8 OB "^ e =° E ST. £ e Si h CB 9 _, i s a s .2 5 ^ c I- fa L a 3 °33 ■? E O CO a w uj S II ■a a. CO 3 T s tf) -C a R H Bj a u t 5 y u II N o a. * o c bo 10 Figure 5. — Porsal view of skull of Auxis rochei (ca. 27 cm SL) on the left and A. thazard (ca. 30 cm SL) on the right. Note that the temporal crests of the A. rochei skull are parallel with each other and with the supraoccipital crest, whereas in A. thazard the temporal crests diverge anteriorly so that the three crests are not distinctly parallel (Yoshida and Nakamura 1965). 40 38 36 32 O AUXIS THAZARD A AUXIS ROCHEI o o o o <* ° o A A ^A 4 i Figure G. — Relation between the width and length of the cranium of Auxis thazard and A. rochei. Width: distance between widest points of pterotic processes; length: distance from anterior margin of vomer to concave ventral tip of aperature of myodome. (Yoshida and Nakamura 1965.) 42 44 46 48 LENGTH OF CRANIUM (mm) 50 Figure 7. — Lateral view of skleton of Auxis thazard (top) and A. rochei (bottom). Note that in A. thazard, the anterior branch of the haemal processes are much longer than in A. rochei and that they are in contact with the preceding haemal arches on the 24th to the 28th vertebrae (Yoshida and Nakamura 1965). 11 2 DISTRIBUTION 2.1 Total area The genus Auxis is distributed worldwide in tropical and subtropical waters.. The confusion surrounding the identification of the two species of Auxis is reflected in their reported distribution in the world's oceans. Fitch and Roedel (1963) concluded from their study that A. rochei was cosmopolitan in distribution whereas A. t hazard was restricted to the Pacific. Collette and Gibbs (1963a) expressed uncertainty about the presence of both species of Auxis in the Atlantic but stated that both probably do occur there. Richards and Randall (1967) independently confirmed that adult A. thazard do occur in the Atlantic. The distribution of A. thazard cannot be separated from that of A. rochei at the present time because of dif- ficulties in the past in distinguishing one species from the other (Yabe et al. 1963). Therefore, the reported dis- tribution in the literature for either A. thazard or A. rochei is questionable and needs to be critically reex- amined. The following discussion takes into account the distribution of both species. Figure 8 shows the distri- bution of Auxis adults in relation to water and land areas. In the Pacific Ocean, Auxis occur off the coast of the United States between Santa Catalina Island and San Clemente Island off California southward into the east- ern tropical Pacific extending as far south as lat. 18°S (Radovich 1961; A. Ch. de Vildoso 5 ). They have also been reported from the Hawaiian (Matsumoto 1960a) and Marquesas Islands (Nakamura and Matsumoto 1967). In the western Pacific, they occur off the coast of Japan as far north as Hokkaido, off Korea, off the coast A. Ch. de Vildoso, Institute del Mar. Callao, Peru, pers. commun. December 1975. of China mainland near southern Manchuria and Ning- po, around Formosa, and in waters surrounding the Ryukyu and Bonin Islands (Rosa 1950). Southward, Auxis have been reported from Samoa Islands and Papua New Guinea (Jordan and Seale 1906), the Philip- pine Islands (Herre 1953), along the eastern and south- ern coasts of Australia (Scott 1962; Laevastu and Rosa 1963; Whitley 1964), Tasmania (Lord 1927), and New Zealand (Roberts et al. 1977). The usual latitudinal range reported for Auxis in the tropical and subtropical waters of the Atlantic Ocean is from lat. 45 °N to 35 °S (Collignon see footnote 2; Miyake and Hayasi 1972). In the eastern Atlantic, they occur infrequently as far north as Bergen, Norway, and the canal of Oslo Harbor (Rosa 1950) and in waters around the British Isles (Went 1955, 1956, 1958, 1967; Rae 1963; Went and Kennedy 1969; Wheeler and Blacker 1969). Auxis also occur in the Mediterranean and Black Seas. Southward, they are found in waters off the Republic of South Africa and offshore around Ascension and St. Helena Islands. In the western Atlantic, the northern- most occurrence of Auxis is off Barnstable, Mass., in the Gulf of Maine (Mather and Gibbs 1957). Southward in the western Atlantic, Auxis have been reported from the Gulf of Mexico, the Caribbean Sea, and the Atlan tic as far south as Mar del Plata, Argentina (Lopez 1961). In the eastern Indian Ocean, Auxis have been reported from the east coast of India and southward along the Indonesian Archipelago to Cape Leeuwin near the southern tip of Western Australia (Rosa 1950; Jones and Silas 1964; Nair et al. 1970). The western Indian Ocean distribution of Auxis extends from the west coast of India offshore to the Maldive and Laccadive Islands, the coast of Iraq in the Persian Gulf (Mahdi 1971), the Red Sea (Ben-Tuvia 1968), and from the Gulf of Aden southward to the coast of Natal in the Republic of South Africa (Williams 1963; Nair et al. 1970). r Y?\ ■v y.y X ~- 1 U V s y r $ . ^ iSS? ^Ci (S> /-■ J ! --■ v * ■^Cx y^ ! i 1 1 j jT yibss*a "^ffiw/jft ^ | '/ 1 ! i ^ 1 A I 1 1 i ' 1 fc 1 tofm k ( ' I 1 1 f i&£$&%K | Y A 20* y/\ tfy. I C? 1&. / 1 1 i i 1 1 1 \ wZw , '4%& i | | '^mC 1 ~< ?V_ '-^ '•// '- ' 1 1 i *w" ►J „, 1 ! \ ! 1 ! 1 ■WKI K | 1 0i 1 v k 1 1 •J i ! ! 2 T tgjgj; 1 " 1 1 \ ! r 1 ! w 1 t •\ ek ■> 1 i -^ r <»' \ t ) W * K « 4{ 1 V. > W » ■ ' ! T / f ^J / 1 i ' _, ! J f V \ J 'J V vj y ' 5 7 % V 4 y My / - s 25 V \i | 1 p 4 0* 6 0* 8 V ' IC c IS V 14 o- 16 0" ie 0* 16 O' \" 2^ ,z °° ,100- e 0* T 6 V 4 y 2 0* < r Figure 8.— The distribution of Auxis spp. adults in relation to water and land areas in the world's ocean. 12 2.2 Differential distribution 2.21 Spawn, larvae, and juveniles The larvae of Auxis, like the eggs which precede them, are displaced from the area of spawning due to drift of the ocean currents (Matsumoto 1958, 1959). Except in a few areas where currents are swift, the actual displace- ment of eggs and larvae appears to be relatively insig- nificant. Whereas the distribution of adult Auxis is usually as- sociated with land masses, that of the larvae has been described as not only coastal but also oceanic. Figure 9 shows the localities of capture of Auxis larvae in the world's oceans. The actual differences in larval and adult distributions, however, may not be real. The adults are usually reported to occur in coastal waters because most of the fishing is done there. But plankton hauls conducted in waters far from land masses have shown that larval Auxis occur in oceanic as well as coastal waters. Matsumoto (1958, 1959) suggested that the localities where larvae of about 3 mm occur prob- ably represent actual spawning sites; therefore, it can be expected that adult Auxis also occur in the oceanic regions of the world's oceans. Watanabe (1964), on the other hand, studying tuna and billfish stomach con- tents, concluded that Auxis are coastal dwellers. The distribution of juvenile Auxis, 10-20 cm SL (stan- dard length), agrees well with that of the larvae, es- pecially near land masses, but in the oceanic regions the presence of juveniles have not been well documented (Fig. 10). Yabe et al. (1963) pointed out that the usual method of using a midwater trawl to collect juveniles has not been successful quantitatively; rather, more in- formation can be obtained through examination of stomach contents of large tunas and billfishes. That juvenile Auxis occur mostly in waters close to land masses is brought out in Table 7 which shows the number of juvenile Auxis taken by midwater trawl in Hawaiian waters. Higgins (1970) observed that in July- September 1967, the catch rate of juvenile Auxis reached 3.5 individuals/tow, the highest among all the 20° 40° 60° 80° 100° 120° 140° 160° 180° 160° 140° 120° 100° 60° 40° 20° 0° 20° 20° 40° 60° 80° 100° 120° 140° 160° 160° 160° 140° 120° 100° 80° 60° 40° 20° 0° 20° Figure 9.— Localities of capture of larval Auxis (Yabe et al. 1963). Figure 10.— Localities of capture of juvenile (10-20 cm SL) Auxis (Yabe et al. 1963). 13 Table 7. — Total number of juvenile Auxis and (in parentheses) the average number per tow collected by a 12 m wide by 8 m high mid- water trawl during the day (1200-1800), night (2000-0200), and morn- ing (0400-1000), RV Townsend Cromwell cruise 32, 12 July-25 September 1967. The trawl, fished at a depth of 100 m during deep hauls and 20 m during shallow hauls, was towed at a speed of 1.5 m/s (Higgins 1970). Area and Number month of tows Frigate tuna Total tunas Inshore Oahu: July 18 7 (0.4) 55 (3.0) August 10 9 (0.9) 28 (2.8) September 8 111 (13.9) 174 (21.8) Total 36 127 (3.5) 257 (7.1) Offshore Oahu: July 11 6 (0.5) 155 (14.1) August 18 380 (21.2) Total 29 6 (0.2) 535 (18.4) Molokai 1 : August 2 1 (0.5) Lanai : September 4 Hawaii: September 12 202 (16.8) Grand total 83 133 (1.6) 995 (12.0) 'The duration of these tows was <6 h. young tunas collected in inshore waters off the island of Oahu. Offshore, the catch rate dropped to 0.2 juvenile/ tow. 2.22 Adults Despite the argument that the distribution of A. thazard cannot be separated from that of A. rochei because of difficulties in distinguishing the two species, a study of stomach contents of tunas and billfishes indi- cated that there are conspicuous differences between the pattern of occurrence of these two species. Watanabe (1964) found that in the Banda Sea, specimens of A. rochei were found in the stomach contents more fre- quently than those of A. thazard, although both species occurred there. This pattern of occurrence reflected the difference in abundance of the two species in the Banda Sea. Off the Queensland coast in Australia, all the specimens collected from tuna and billfish stomachs were A. rochei. 2.3 Determinants of distribution changes The extreme southern boundary of the distribution of Auxis in the Indian Ocean lies at about lat. 36°S which is extremely close to the position of the 20°C isotherm for the greater part of the year, including the southern summer (Williams 1963). Off South Africa and Australia, the occurrence of Auxis coincides with the time of maximum water temperature. Likewise, off East Africa and the Seychelles, Auxis occur in the months of the northwest monsoon when the temperature is max- imal at about 29°-30°C. The appearance of Auxis in East African inshore waters also coincides with the time of greatest fertility of the surface waters. Temperature, it has been shown, is clearly a highly important variable in explaining the distribution of Auxis larvae. Klawe et al. (1970) observed that the op- timum temperature of surface waters under which larval Auxis are found is between 27.0° and 27.9°C (Fig. 11). Their tolerance for temperature, however, is very wide. Richards and Simmons (1971) determined that Auxis larvae occur in waters with surface temperatures as low as 21.6°C and as high as 30.5°C, the widest range among any of the tuna larvae they studied. The vertical distribution of larval Auxis has been reported to be limited to the layer above the thermo- cline. Comparing average catches of tuna (including Auxis) larvae for two types of plankton tows made in the eastern Pacific, Klawe (1963) found that the surface tows usually caught 9.2 times as many larvae as the deep tows. But comparison of catches made by surface and by 140 m oblique tows showed that the former caught an average of only 3.2 times as many larvae as the latter. In fact, the numbers of larvae caught in the simultaneous surface and oblique tows were signifi- cantly correlated. Klawe found that the relative number of tuna larvae taken in surface and oblique hauls ap- proximates the ratio obtained from the depth of the obli- que tow to the depth of the layer above the thermocline thus substantiating the conclusion earlier reached by others that larvae are limited to the layer above the thermocline. Among other determinants of larval distribution that have been examined are salinity, plankton, and light conditions. Klawe et al. (1970) found no relationship be- tween zooplankton volumes and larval catches; therefore, their findings are in agreement with those of Strasburg (1960) and Nakamura and Matsumoto (1967) who observed this phenomenon for tuna larvae in - 0.5- 19 20.0 210 22 230 24.0 25 26 27 280 29.0 I I I I I I I I I I I 199 20.9 21.9 229 239 249 25.9 26.9 27 9 289 299 TEMPERATURE (°C) Figure 11. — Average catches, log 1 " (number caught +1), of larval Auxis grouped according to surface temperatures at the time of capture (Klawe et al. 1970). 14 general. Salinity, likewise, showed no relationship with catches of larval Auxis, but Klawe et al. (1970) reasoned that this lack of relationship may be due to the narrow range of salinity encountered during the period that the study was made. Salinity, therefore, cannot be dis- counted as a factor affecting distribution. In fact, Richards and Simmons (1971) determined that larval Auxis were found in a relatively narrow salinity range from 33.2%. to 35.47.. compared to that of other tunas, with the excep- tion of little tunny, Euthynnus alletteratus. Concerning the effects of light intensity, Higgins (1970) observed that juvenile Auxis were more vulnerable to the midwater trawl at night than during daylight hours. Furthermore, they were more readily caught in shallower than in deeper tows, although the data suggested that they may occur as deep as 100 m. Strasburg (1960) reported that larval Auxis occur at the surface very infrequently during daylight hours but are often found there in large numbers at night. Noting that this pattern of an increase in catch at night could result from either net avoidance or vertical migration, Stras- burg reasoned that if net avoidance only were involved, one would expect the catch to be essentially constant at night. The results of his study proved otherwise; there- fore, he concluded that vertical migration appears to be the major factor causing the increase in surface catch at night. Contrarily, in a later study by Klawe et al. (1970), it was observed that larval Auxis show no significant diel movement in the water column. From a statistical test they conducted for time of day as well as for inter- action between time of day and type of tow, Klawe et al. (1970) found no interaction between type of tow and time of day; therefore, they concluded that no vertical movement had occurred. They reasoned that any ability larval Auxis may have had to avoid the sampling gear was independent of time of day. 2.4 Hybridization 2.41 Hybrids; frequency of hybridication; species with which hybridization occurs; methods of hybridization There has been considerable progress in studies on egg and larval development of tuna and tunalike fishes in recent years. Most impressive are the experiments by Japanese scientists on artificial fertilization of eggs and the subsequent rearing of the larvae. Among the tuna species that have been used in these experiments were yellowfin, skipjack, frigate, and bullet. In 1970, the Japan Fisheries Agency coordinated a 3-yr cooperative program to artificially fertilize and rear tunas (Ueyanagi et al. 1973). The experiments were conducted through the cooperative efforts of Tokai University, Kinki University, Shizuoka Prefectural Fisheries Experimen- tal Station, Mie Prefectural Owase Fisheries Experi- mental Station, Nagasaki Prefectural Fisheries Experi- mental Station, and the Far Seas Fisheries Research Laboratory of the Japan Fisheries Agency. Experiments conducted at Kinki University Fish- eries Experimental Station (1974) involved cross- fertilization of A. thazard and A. rochei. The ripe eggs taken from A. thazard caught by set net at Kashino, Oshima, on 17 July 1973 at 0530 were fertilized and transported to the experimental station at Shirahama. At 1350 on the same day, there were about 35,120 viable eggs which averaged 0.86 mm in diameter and about 7,000 dead eggs. Twenty-five thousand viable eggs were placed in a 3-ton (3,000 liter) tank and 10,000 in a 1-ton (1,000 liter) tank. Water temperature in the tanks was about 26°C. Larvae started to emerge the next day, approximately 30 h after fertilization and all hatching was completed during the day. The larvae were kept in seawater with Chlorella which had a density of 400,000/cc. Aeration was also provided. Rotifers were added to the tank 2 days after hatching and marine plankton was added the next day. All the larvae in the 3- ton tank died after 6 days. In the 1-ton tank, although mortality was heavy, there were some survivors. At 14 days, the water was circulated, and at 18 days after hatching, artemia and eggs of frozen yellowtail, Seriola quinqueradiata, were used to supplement the marine plankton. Mortality continued and the last survivor, which attained a length of 4.8 cm and a weight of 0.907 g died after 31 days. 3 BIONOMICS AND LIFE HISTORY 3.1 Reproduction 3.11 Sexuality Like all other scombrids, Auxis are heterosexual. There are no externally visible characters that aid in distinguishing males from females. Internally, the paired and elongated gonads are nearly symmetrical and are suspended by mesenteries extending almost the entire length along the roof of the abdominal cavity. At the posterior end of the abdominal cavity, the gonads extend along both sides of the anal fin. This posterior ex- tension, according to Kishinouye (1923), is due to the narrowness of the abdominal cavity. 3.12 Maturity Studies of gonads of troll-caught fish from waters of Kaneohe Bay, Oahu, in the Hawaiian Islands indicate that frigate tuna are nearly mature at a size of about 35 cm (Tester and Nakamura 1957), but in Japanese waters, they have been reported to reach maturity at about 29 cm. Yasui (1975) investigated the use of liver weight as a possible index of sexual maturity. He plotted liver weight against body length of frigate tuna caught off the Izu Islands and off Mera, Shizuoka Prefecture, and observed a point of discontinuity at a body length of 29 cm. His data also showed that 97% of the fish <29 cm were caught after September whereas 95% of those larger than that were caught before August. The repro- ductive index, according to Yasui, was highest in July. 15 Observing that almost all fish >29 cm had high repro- ductive indices, Yasui hypothesized that liver weight is related to oogenesis and that it can be used as an indi- cator of sexual maturity. In the Spanish trap net fishery located in and around the Strait of Gibraltar, the size of A. rochei at first spawning is 35 cm for the females and 36.5 cm for the males (Rodriguez-Roda 1966). Rodriguez-Roda classi- fied the testes and ovaries according to the develop- mental stages as follows: I Immature II Maturing III Ripening IV Prespawning V Spawning VI Spent Rodriguez-Roda noted that in May, a large proportion of the males and females were in stage IE. In June- August some showed development in stage IV, but by September about one-third of the males and one-fourth of the females sampled had gonads that appeared to be spent (Table 8). Rodriguez-Roda noted that gonad clas- sification based on gross characteristics such as he used could lead to misleading conclusions; however, he indi- cated that the relative gonad weight (gonad weight in relation to body weight) tended to confirm his conclu- sions. 125 g, respectively. According to Rao, the smaller of the two fish may "have been captured while in the act of spawning since most of the ripe ova were already lost by the time it was examined." 3.16 Spawning Earlier, it was pointed out that although the plank- tonic eggs and larvae of Auxis become displaced from the area of spawning due to ocean currents, the dis- placement is insignificant. Therefore, localities where larvae of 3 mm or less occur probably represent actual spawning sites. In the eastern tropical Pacific, Auxis larvae are the most abundant among all the scombrid larvae collected. Ahlstrom (1971) identified 1,563 out of 1,919 scombrid larvae in the EASTROPAC collection as Auxis spp. Based on this information and the distribution of the catches of larval Auxis, Klawe (1963) observed that off Baja California, spawning occurred in coastal waters whereas to the south it appeared to be in more oceanic waters away from continents or islands. The most north- erly area of spawning appeared to be near Cedros Island and at the head of the Gulf of California. To the south, Klawe found the limit of spawning to be off Point Santa Elena in Ecuador. He delimited the general area of Table 8. — Percentages of male and female Auxis rochei from the Spanish trap net fishery classified as I - immature, II - maturing, III - ripening, IV - prespawning, V - spawning, and VI - spent in May-September in 1958, 1961, 1963, and 1964 combined (Rodriguez-Roda 1966). Males Females Month I II III IV V VI N I n ra IV V VI N May 8.57 88.57 1.43 1.43 70 12.86 75.71 — 11.43 62 June — 2.94 79.41 17.65 — — 34 — 6.06 69.70 24.24 — — 33 July — — 90.00 10.00 — — 30 — 20.00 80.00 — — — 20 August — 2.74 97.26 — — — 73 — 1.35 94.59 2.70 — 1.35 74 September 20.00 14.29 25.71 5.71 1.48 32.86 70 277 32.61 17.89 15.22 8.70 26.09 46 235 3.13 Mating No information is available on the mating habits of Auxis but, in general, it is believed that scombrids release their sexual products directly into the water without pairing of the male and female. The mechanism which triggers the reproductive activity, i.e., spawning and fertilization, is still unknown. 3.14 Fertilization Fertilization is external. 3.15 Gonads The gonads of Auxis are paired, elongate organs suspended from the dorsal wall of the body cavity by lengthwise mesenteries. Rao (1964) described the ovaries of A. thazard in "spawn-ripe" condition as pink- ish-pale yellow organs. Two samples of ovaries coming from fish measuring 41.6 and 44.2 cm weighed 52 and occurrence by drawing a straight line from Point Santa Elena to the intersection of lat. 10°N and long. 116°W and from there to Cedros Island off Baja California. In the extreme eastern Pacific, A. thazard spawn throughout the year, although in the northern region, it may be restricted (Figs. 12, 13). This phenomenon of seasonal spawning in the northern region of the eastern Pacific may be similarly reflected in the most southern region, but data are inadequate to confirm this. Data from collections made off Cape Blanco indicated that Auxis spawn off Costa Rica throughout the year but peak spawning occurs in December- April (Table 9). Klawe et al. (1970), however, observed that the period of peak spawning of A. thazard may vary, because their data suggested that larvae were more abundant in August-November with a peak in October (Fig. 14). They suggested that differences in spawning peak between their samples and those examined earlier by Klawe (1963) may be due to differences in sampling locations within the general area, to environmental changes, or other unknown causes. 16 Figure 12. — Relative abundance (number of larvae per 1,000 m 3 of water strained at the surface) of Auxis in May-October, based on plankton collections made in the eastern tropical Pacific. Dashed lines enclose area under investigation by Inter-American Tropical Tuna Commission in 1952-59; black dots indicate positions of plankton hauls with zero catches. (Klawe 1963.) Table 9. — Average number of Auxia larvae caught per hour during different months off Cape Blanco, Costa Rica, with aim (silk or nylon) net towed at 1.0 m/x (Klawe 1963). Month Number of Month of Number of capture larvae capture larvae January 7.6 July 0.2 February 1.0 August 0.9 March 21.7 September 3.5 April 18.3 October 0.1 May 3.8 November 1.3 June 0.5 December 6.0 From the centred Pacific, there is evidence that Auxis spawn not only in waters around the Hawaiian Islands but also to the west and south of these islands. The pres- ence of small larvae of 3 mm or less in plankton samples indicated that spawning probably occurred in these waters just a few days prior to sampling (Table 10) (Matsumoto 1958; Strasburg 1959). Further evidence of spawning has been reported by Yoshida and Nakamura (1965), who observed that among the fish landed from schools of A. thazard and A. rochei fished by pole and line off Kaena Point, Oahu, milt was flowing from the vent of males of both species. The females, however, showed no signs of free-flowing eggs even after pressure was applied to the abdomen externally. Subsequent ex- amination revealed that the females had ovaries that were flaccid, translucent, and grayish such as those usually seen in spent females of yellowfin tuna (June 1953). 17 I I ) I I I r- I I I I i I I I I I I I i I I i —r- i i i i i i i i 115° 110° 105° 100° 95° 90 I i i eg! Figure 13.— Relative abundance (number of larvae per 1,000 m 3 of water strained at the surface) of Auxis in November-April, based on plankton collections made in the eastern tropical Pacific. Dashed lines enclose area under investigation by Inter-American Tropical Tuna Commission in 1952-59; black dots indicate positions of plankton hauls with zero catches. (Klawe 1963.) OCT. NOV. DEC. JAN. FEB. APR. AUG. Figure 14. — Average monthly catches, log l0 (number caught =1), of larva Auxis during cruises in October-February, April, and August to the entrance of the Gulf of California (Klawe et al. 1970). Auxis have also been reported to spawn in waters around the Marquesas Islands (Nakamura and Matsu- moto 1967). From samples of tuna larvae collected at the diel variability station (station occupied for 24-h periods), Nakamura and Matsumoto found that skip- jack tuna, Katsuwonus pelamis, larvae were most numerous followed by Auxis larvae. Because currents around the Marquesas Islands are suspected to be weak, Auxis larvae as well as those of other tunas could not have drifted very far from the spawning site; therefore, 18 Table 10.— Number and size range of larval Auxis thazard (3 mm or less) collected in 1-h horizontal hauls (three 1 m nets towed simultaneously at different depths) and 30-min oblique hauls (surface to 200 m) by the RV Hugh M. Smith in Hawaiian and equatorial waters (Matsumoto 1958). Cruise and Time tow Depth of Number of Size range station number Date started haul (m) larvae (mm) Hawaiian waters: Cruise 6: No. 1A 25 Aug. 1950 1405 0, 50, 150 1 3.6 No. 7 27 Aug. 1950 1853 ■ 0, 50, 150 1 6.2 No. 8 28 Aug. 1950 0507 0, 100, 200 4 2.5-8.2 No. 10 24 Aug. 1950 2239 0, 150, 300 10 2.7-3.7 No. 11 25 Aug. 1950 0500 0, 50, 150 2 7.02,8.1 No. 13 24 Aug. 1950 0528 0, 50, 150 12 3.0-5.2 Total 30 2.5-8.2 Equatorial waters: Cruise 5: No. 21 9 July 1950 0143 0-200 2 2.6 No. 51 5 Aug. 1950 2300 0-200 2 4.2,5.5 Cruise 14: No. 26 1 Feb. 1952 1920 0-200 1 3.6 No. 29 19 Feb. 1952 2008 0-200 1 3.1 Cruise 15: No. 3 29 May 1952 0315 0-200 1 7.2 Cruise 18: No. 31 9 Nov. 1952 2329 0-200 1 4.0 Total 8 2.6-7.2 Figure 15. — Localities of capture of larval and postlarval Auxis in Philippine waters (wade 1951). it is reasonable to assume that Auxis spawning occurs throughout Marquesan waters. The western Pacific also has a number of probable spawning areas. Figure 15 shows the locations where lar- val and postlarval Auxis (4.2-11.6 mm) were captured in plankton tows in Philippine waters. Wade (1951) noted that Auxis larvae were most abundant in January- March whereas they were scarce the remainder of the year. Furthermore, by examining the time of capture of juvenile Auxis in fish traps and nets close to shore, Wade determined that small juveniles first appear on the market in late November, become abundant in January-February, then become scarce so that after May there were none. From these observations, he con- cluded that adult Auxis spawn in protected, more or less shallow areas, fairly close to land, that the main spawn- ing season is during the winter months, and that scattered spawning takes place the remainder of the year. Other western Pacific spawning sites include waters around Australia and Japan and the Celebes and South China Seas. The Tasman Sea off Sydney, Australia, where larval Auxis (sizes not given) have been captured, is reported to be a possible spawning site (Fig. 16). Spawning also appears to be of significant intensity in the Gulf of Tonkin (Gorbunova 1965b). Whitley (1964) reported that larval and juvenile Auxis are common in the Tasman Sea. Two types of Auxis larvae were iden- tified from the collection made by the Dana Expedition in 1929 (Matsumoto 1959, append, table 2); these specimens probably represent A. thazard and A. rochei. According to Whitley (1964), similar specimens were washed up on Collaroy Beach in August- September . 'f v r^ PHILIPPINE" vM>V ISLANDS 1958. The presence of larval and juvenile Auxis in waters of the Tasman Sea would indicate that adult Auxis also spawn in this region of the Pacific. In Japanese waters, A. rochei with ripening gonads are caught in April-June, whereas those with ripe and 19 20 spent gonads were taken in June-July and July-August, respectively (Suzuki and Morio 1957). This was con- firmed by Hamada, Morita, Ishida, and Takezawa (1973), who examined the gonad condition factors of A. rochei caught in waters off Kochi Prefecture and sug- gested that spawning took place in late June. They also examined the gonad condition of A. rochei taken off Taiwan in June-July and found that the gonad index was very high indicating perhaps that spawning was im- minent. Concerning the larvae, Yabe and Ueyanagi (1962) collected those of A. rochei from May through July near Japan whereas Yokota et al. (1961) obtained larval A. rochei, measuring 3-5 mm TL (total length), from south of Shikoku and Kyushu in June-August but most were taken in July. Farther south, however, in the Celebes and South China Seas, larval A. rochei were col- lected predominantly during January and February (Yabe and Ueyanagi 1962). It has also been established that Auxis spawn in the Indian Ocean (Bogorov and Rass 1961) and that spawn- ing extends for at least 8 months from August to April; however, there is no information on their spawning ac- tivity for the rest of the year. Among the tuna larvae col- lected by the Dana Expedition from south of the equator between lat. 3° and 24°S at long. 50°E, Jones and Kumaran (1963) found 131 larval A. thazard, indicat- ing that this species spawned in this area in December- January. North of the equator in the Laccadive Sea, they also found larval A. thazard in the samples indi- cating a spawning period in January-April, slightly later in the year than in the south. Rao (1964) added that among samples of A. thazard he collected from commer- cial landings, the percentage of maturing and mature ovaries was highest in March. In August and September, the majority of the fishes examined were ac- tually in spawning condition and some individuals com- pletely spent. Samples obtained subsequently in November contained only spent fish. Although evidence indicates that A. thazard spawn in the Indian Ocean, present data concerning spawning of A. rochei are still too meager to draw definite conclu- sions. Jones and Kumaran (1963) found 20 larval A. rochei in the Dana collection from the Indian Ocean. All were collected in December-January from three loca- tions between Madagascar and the African coast. Mat- sumoto (1959), who examined the collection made by the Dana from the eastern Indian Ocean southwest of the Sunda Archipelago in August-November, recog- nized the presence of two types of Auxis larvae. Gor- bunova (1963) confirmed Matsumoto's observations, reporting that two types of Auxis larvae were distin- guished in the Soviet collection made in the Indian Ocean. From the information available, Jones and Kumaran (1963) hypothesized that low latitudinal areas of the eastern Indian Ocean southwest of the Sunda Archipelago and the western Indian Ocean between Madagascar and the coast of Africa are possible spawn- ing grounds of A. rochei. Gorbunova (1965a, 1965c) col- lected A. thazard larvae of 5-11 mm in February and March in the Bay of Bengal. In the Atlantic Ocean, tuna larvae have been col- lected off the west coast of Africa, mainly in the region of Dakar at the western tip of Africa, and off Takoradi in the Gulf of Guinea (Kazanova 1962; Richards 1969). Among the various species of larval tuna and tunalike fishes collected, those identified as Auxis were most numerous. Kazanova had at his disposal a series of Auxis larvae some as small as 3 mm. Furthermore, in addition to larvae identified as A. thazard, there were others which were similar to A. thazard but differing in some characteristics. Kazanova referred to them as the second of the two groups (Type II of Matsumoto 1959) of Auxis larvae. Therefore, it can be inferred that these regions of the Atlantic are also spawning grounds of the two species of Auxis. Certain areas in the Mediterranean Sea have also been suggested as possible spawning sites of Auxis sp. Ehrenbaum (1924) reported Auxis spawning from July to September in the Mediterranean. Off Greece, spawn- ing of Auxis was also reported to occur from June to September and in the Gulf of Catania (Sicily) in June and July (Belloc 1954). Declerc et al. (1973) reported capturing recently hatched larvae of Auxis in waters surrounding the Balearic Islands, located off the Mediterranean Spanish coast. In August-September 1971, 140 larval Auxis were captured by the RV Ichthys and in June-July 1972, an additional 52 Auxis larvae were taken. According to the authors, larval Auxis was the most abundant among the tuna larvae captured which included Thunnus thynnus, T. alalunga, and Sarda sp. The presence of the planktonic stages of all these tuna species in the plankton-net catches provided evidence that these species spawned in waters near the Balearic Islands. The waters around Tunisia have yielded Auxis in a mature state, and this bit of evidence has led Postel (1964) to believe that spawning probably takes place in Tunisian waters. Postel examined the gonads of Auxis caught in madragues (traps) off Tunisia, found them mature, and concluded that spawning occurs in June- August. For Auxis that were taken off Algeria, Postel also determined that they spawn there probably in August. Off the northwestern coast of Africa, Cape Verde Islands and Dakar have been mentioned as spawning sites of Auxis (Frade and Postel 1955; Kazanova 1962). Gorbunova (1965c) reported capturing about 600 Auxis larvae at one station in the Gulf of Aden. In the western Atlantic, the Gulf of Mexico appears to be a large spawning ground for Auxis. Hayasi (1972) reported collecting tuna larvae identified as yellowfin, bigeye, skipjack, and Auxis in gulf waters (Fig. 17). Klawe and Shimada (1959), who collected juvenile Auxis in March-April and in June-August from a large number of stations spread widely over the Gulf of Mex- ico, stated that the north central gulf region is definitely a spawning locality for this species. The area southeast of the Mississippi River Delta is another possible spawn- ing site because most of the larvae were captured there, but Klawe and Shimada pointed out that the collection 21 Figure 17.— Plankton stations occupied during cruises of the RV Shoyo Maru in 1969-70. Solid triangles represent stations at which Auxis larvae were collected. (Hayasi 1972.) may reflect sampling intensity rather than actual dis- tribution of young Auxis. In waters adjacent to the gulf extending from Cape Hatteras to Cuba and particularly in the Strait of Florida, larval Auxis have shown up in collections made in February-March (Tibbo and Beckett 1972) and in May-June and in August (Klawe 1961) indicating the possibility that spawning occurred during these months in this region. 3.17 Spawn The eggs of Auxis are pelagic. Rao (1964), who ex- amined ovaries of A. thazard from the Indian Ocean, observed that ripe eggs flowed freely out of the ovaries when slight pressure was applied on the abdomen. Fresh ripe eggs were perfectly spherical, had a colorless homogeneous yolk mass, and had an average diameter of 0.97 mm (range of 0.88-1.09 mm). Preserved eggs were somewhat translucent and had an average diameter of 0.86 mm (range 0.78-0.98 mm). Each ripe egg contained a fairly large, spherical, single oil globule which averaged 0.22 mm (range 0.21-0.22 mm) in diameter. Like other scombrids, female Auxis do not extrude all their ripe eggs during spawning. Microscopic examina- tion by Yoshida and Nakamura (1965) of ovaries from eight A. thazard and five A. rochei caught in Hawaiian waters revealed that both species had eggs in various stages of resorption. One specimen of A. thazard still contained residual eggs in the lumen; their diameters varied between 0.75 and 1.30 mm and averaged 1.08 mm. No free residual eggs were found in A. rochei. Concerning primordial eggs, two size groups — one rang- ing from 0.07 to 0.10 mm and the other from 0.17 to 0.22 mm in diameter — were noted in ovaries of A. thazard. In A. rochei, there was only one group of primordial eggs ranging from 0.07 to 0.08 mm in diameter. The development of eggs and larvae of Auxis has been described for the two species from the Atlantic and Pacific Oceans. Mayo (1973) successfully hatched eggs of two species of Auxis collected from the Straits of Florida and described the growth, behavior, ecology, and development of the larvae. In the Pacific, artificial fertilization of eggs and subsequent rearing of larvae were conducted by Japanese scientists and their results have been briefly discussed in section 2.41. Harada, Murata, and Miyashita (1973) and Ueyanagi et al. (1973) published the results of these tuna-rearing ex- periments in which fish in advanced stages of sexual maturity were used. Of three conventional fishing methods attempted, only purse seining proved effective in capturing mature yellowfin and skipjack tunas and set nets for collecting mature Auxis. The "dry method" was used to obtain mature eggs and sperm. Mature eggs from ripe female Auxis general- ly oozed out when pressure was applied to the abdomen; therefore, no incision was necessary (Ueyanagi et al. 1973). The investigators then obtained mature males and applied pressure on the abdomen to force the sperm-bearing milt over the eggs. The mixture was gent- ly stirred before any water was added. Collected and fertilized at sea, the eggs were then transported to various laboratories to be hatched (Ueyanagi et al. 1973). Fertilized eggs of Auxis were transported to laboratories in 2-4 h on eight different oc- casions. The mortality of the eggs, which were trans- ported in plastic bags filled with oxygen-saturated seawater, was negligible. The details of the success obtained in fertilizing and rearing both A. thazard and A. rochei follow. Auxis thazard Mature eggs of A. thazard caught at Mera, Shizuoka Prefecture, were collected, fertilized, and successfully hatched on five occasions between 15 July and 15 August 1971 (Ueyanagi et al. 1973). Larval rearing was most successful on the first trial when survival lasted up to 10 days. Newly hatched larvae emerged 25 h after fer- tilization and measured 2.59 mm TL. Heavy mortality usually occurred 4-5 days after hatching. At Kushimoto, Wakayama Prefecture, mature eggs were collected, fertilized, and transported to the Marine Laboratory of Kinki University on three occasions in 1972 (Harada, Murata, and Miyashita 1973; Ueyanagi et al. 1973). The spherical, fertilized eggs were buoyant and varied from 0.93 to 0.98 mm in diameter (Fig. 18). Emerging 34-62 h after fertilization at temperature ranging between 21.4° and 23.5°C, the newly hatched larvae measured from 3.26 to 3.60 mm TL. On the third day after hatching, the larvae were fed rotifers, which were replaced by marine plankton, mainly copepods, on the seventh day. Twenty days after hatching, the larvae were fed several types of fish flesh. The larvae were placed in various-sized aquaria to determine the effects of tank capacity on growth and survival; the investi- gators found that growth was rapid in the 0.5-ton (500 liter) aquarium during the first 30 days after which time the rate slowed appreciably (Table 11, Fig. 19). Max- imum survival was 36 days. Larvae reared in aquaria which had capacities of 3 tons (3,000 liters) and 70 tons 22 <"> •v Fertilized egg of .4 u.w (hazard (hirasQda) in morula stage. Q.95mm in diameter. 20 hours after fertilization. a Postlarva about 5 days old. Postlarva about 10 days old. iHH!M'II|fMHi''l!HO'HfHlf«H»m»**»^"*"»^ Just before hatching, 40 hours. 2 3 t 5 6 . Young fish about I 7 days old, 64mm in total length. • •■ o "* r 7T' , V il|!i '" u i".': | l | !,'l!! r il l llll|llli|li, T ! l l l U,| Ill)ll,l,ltlfl!!, .. •< 4 SI 6 ! 7 „' . ,l |lg ,l, , ., •>•*»•» * I arva about I day old. 3.4mm in total length. Young fisn aboul 40 days oW , :0mm , n , otJ j lcng(h Figure 18. — Various stages of development of egg and larvae of A uxis thazard (Harada, Murata, and Miyashita 1973). (70,000 liters) grew faster than those reared in the 0.5- ton aquarium, reaching a length of 12 cm in 33 days (Table 12, Fig. 19). Growth decreased significantly after that and the last larvae died on the 41st day. Auxis rochei In 1972, the collection and fertilization of eggs from mature A. rochei succeeded in seven trials at Mera, 23 Table 11. — Measurements on Auxis thazard larvae reared in a small (500 liter) aquarium (Harada, Murata, and Miyashita 1973). Days after hatching; Total length (mm) Fork length (mm) Body depth (mm) ~i 1 1 r~ • • TOTAL LENGTH ' o o WATER TEMP 1 1 TOTAL LENGTH a fl WATER TEMP I — IN LARGE TANKS — IN A SMALL TANK Body weight (g) 10 5.95 — 0.97 — 10 4.90 — 1.10 — 15 11.4 — 2.6 — 20 29.0 25.0 5.0 — 28 55.0 52.0 9.5 — 28 53.5 50.0 9.0 — 30 56.0 53.0 10.0 — 30 47.0 43.0 9.0 — 31 63.0 59.5 10.5 — 31 46.0 43.0 8.5 — 33 55.0 51.0 10.0 1.1 33 55.5 52.0 9.5 1.1 36 60.0 56.5 10.0 1.8 30 Shizuoka Prefecture (Ueyanagi et al. 1973). In one ex- periment, the larvae, which were fed rotifers, copepods, larvae of sea bream and goby, chopped bivalve meat, and chopped fish flesh, survived up to 67 days after fer- tilization. In other rearing trials of A. rochei conducted at the Shirahama Fisheries Laboratory of Kinki University, Wakayama Prefecture, in June 1972, about 40,000 eggs were collected from mature females caught in a trap net near Oshima Island by Cape Shionomisaki, fertilized, and transported to the laboratory (Harada, Murata, and Furutani 1973). The buoyant, spherical fertilized eggs, which measured between 0.97 and 0.99 mm in diameter (Fig. 20), hatched 38-52 h later at temperatures ranging from 21.4° to 23.5°C. The newly hatched larvae ranged from 3.3 to 3.6 mm TL. Fed the same diet as that given larvae of A. thazard as previously described and con- fined in one small 0.5-ton aquarium and two large aquaria of 3- and 30-ton (30,000 liter) capacities, the lar- vae of A. rochei, like those of A. thazard, grew faster in the large aquaria than in the small aquarium (Tables 13, 14; Fig. 21). Table 13.— Measurements on Auxis rochei larvae reared in a small (500 liter) aquarium (Harada, Murata, and Furutani 1973). Days Total Fork Body Body after length length depth weight hatching (mm) (mm) (mm) (g) 3.49 0.62 2 3.58 — 0.70 — 10 5.18 — 1.05 — 15 7.10 — 1.70 — 15 6.65 — 1.15 — 15 8.50 — 1.95 — 15 9.10 — 2.25 — 20 11.00 — 2.70 — 20 9.95 — 2.50 — 20 8.50 — 2.35 — 20 9.50 — 2.60 — 43 95.00 89.00 14.00 7.0 43 89.00 85.00 15.00 6.2 15 20 25 30 35 DAYS AFTER HATCHING Figure 19. — Growth curves of Auxis thazard larvae reared in two large (3,000 and 70,000 liter) and one small (500 liter) aquaria. Length measurements are based on a single fish. (Harada, Mura- Table 14. — Measurements on Auxis rochei larvae reared in two large (3,000 and 30,000 liter) aquaria (Harada, Murata, and Furutani 1973). la, ana ivu yasnua i\)i6.) Days Total Fork Body Body after length length depth weight hatching (mm) (mm) (mm) (g) Table 12. — Measurements on large (3,000 and 70,000 liter) Auxis thazard larvae reared in two aquaria (Harada, Murata, and Miya- 2 3.49 3 58 - 0.62 73 - shita 1973) 18 36 40 51 51 49.00 88.00 140.00 152.00 144.00 82.0 120.0 142.0 138.0 7.70 18.00 30.00 32.00 24.00 0.8 Days after hatching Total length (mm) Fork length (mm) Body depth (mm) Body weight (g) 4.2 25.0 32.5 21.9 3.47 — 0.66 — 51 152.00 143.0 28.00 28.2 2 3.8 — 0.78 — 52 145.00 136.0 26.00 24.0 17 64.0 — 7.5 1.8 52 157.00 152.0 28.00 37.0 33 120.0 112.0 18.0 10.6 52 155.00 145.0 26.00 31.7 33 89.0 84.0 14.0 5.5 58 120.00 112.0 18.00 — 33 120.0 115.0 18.0 10.6 58 155.00 145.0 30.00 — 24 Fertilized eggs of Auxis rochei (marusoda) in morula stage, 0.98 mm in diameter. Postlarva about 5 days old. ,><"?!L*S 20 hours after fertilization. S^ 28 hours after fertilization. V .3:' Just before hatching, 38 hours. i&c- a * Postlarva about ID days old. n nut |i ii!(!H(lt|!tttftf itimiitimtiitntinirimM' Young fish about I 1 ) days old, 49mm in total length. Larva about 1 day old. 3.5mm in total length. Immature fish about 52 days old. 157mm in total length. Figure 20. — Various stages of development of egg and larva of Auxis rochei (Harada, Murata, and Furutani 1973). Experiments on artificial fertilization were continued in 1973. On 15 July 1973, at 0500, ripe eggs were col- lected from A. rochei caught in a set net off Kashino, mental Station 1974). The water temperature at the time of fertilization was 25°C. The fertilized eggs, which averaged 0.903 mm in diameter, were transported to the Oshima (Kushimoto, Wakayama Prefecture), and arti- experimental station at Shirahama. The following day, ficially fertilized (Kinki University Fisheries Experi- an estimated 7,000 newly hatched larvae were placed in 25 1 1 r • • TOTAL LENGTH o o WATER TEMP 16 h * * TOTAL LENGTH A A WATER TEMP IN LARGE TANKS IN A SMALL TANK 30 20 IE 20 30 40 DAYS AFTER HATCHING Figure 21. — Growth curves of Auxis rochei larvae reared in two large (3,000 and 30,000 liter) and one small (500 liter) aquaria. Length measurements are based on a single fish. (Harada, Mura- ta, and Furutani 1973.) a 3-ton aquarium and at age 2 days were fed rotifers and marine plankton. By age 18 days, the larvae were 3 cm long. From the 24th day, the larvae were fed minced fish flesh, and they grew rapidly, reaching 5-6 cm by the 27th day; however, heavy mortality reduced the number of larvae to 36. By the 35th day, the survivors were 7.7-7.9 cm long. Only one larva survived for 50 days. The Japanese investigators observed that this rate of growth was slower than in previous experiments and believed that it was due to lack of fish larvae as food. They em- phasized, however, that the experiments succeeded in artificially producing "seedlings" of A. rochei which were several centimeters in length. In general, then, these experiments on frigate and bul- let tunas have shown that: 1. the eggs are buoyant, spherical, and measure between 0.93 and 0.99 mm in diameter; 2. the eggs hatch from 34 to 62 h after fertilization in water temperatures ranging between 21.4° and 23.5°C; 3. the newly hatched larvae measure 3.26 and 3.60 mm TL; 4. the larvae can be fed rotifers, copepods, and minced fish flesh, in that order, after hatching; 5. the growth and survival rates of larvae are better in large rather than in small aquaria; and that 6. under the best conditions, A. thazard larvae grew 120 mm in 33 days whereas A. rochei larvae grew 157 mm in 52 days. 3.2 Preadult phase 3.21 Embryonic phase See section 3.17 above. 3.22 Larval phase Studies of anatomical development of larval Auxis show that gross nucleation of the central nervous system of young Auxis (3.6-23.0 mm SL) is similar to young Thunnus albacares but heavier than in Katsuwonus pelamis, T. obesus, T. thynnus, T. alalunga, T. atlan- ticus, and Euthynnus alletteratus (Richards and Dove 1971). Using sectioned organs and tissues, Richards and Dove determined that the swim bladder in katsuwonus and Auxis, which, among scombrids is similar in the very early stages of development, starts to degenerate when the larvae reach the size of 9 mm, and is nearly completely degenerated in specimens 24 mm long at which time only the gas gland remains. Richards and Dove also determined that in Auxis the kidney is short until the larvae reach 11 mm after which it lengthens significantly and that the postcardinal vein is larger in Auxis and Euthynnus than in other species. The criteria for specific identification and separation of Auxis larvae from those of other tuna and tunalike fishes appear to be quite well established. Ueyanagi (1964) has provided a key to identifying tuna larvae and this is given below: Key to the larvae of tunas from the western Pacific 8 a Melanophores visible on the forebrain bj One conspicuous melanophore present on the caudal peduncle. No chromato- phores on the isthmus or directly ante- rior to the anus. .... skipjack tuna, Katsuwonus pelamis b 2 There is a dotted line of melanophores along the midlateral line on the poste- rior part of the body. Chromatophores appear on the isthmus and directly an- terior to the anus. kawakawa, Euthynnus af finis a 2 No melanophores visible on the forebrain. c Melanophores appear on the sides of the body, d Chromatophores appear on the isth- mus and directly anterior to the anus. Melanophores appear as dotted lines along the middorsal, midlateral, and midventral lines of the caudal region (usually in three lines). .... frigate and bullet tunas, Auxis spp. d 2 Chromatophores do not appear on the Applicable to larvae under about 8 mm TL. 26 isthmus nor directly anterior to the anus. f l One or several melanophores ap- pear along the dorsal and ventral sides of the body, g The most anterior of the chro- matophores on the dorsal side is located posterior to the origin of the second dor- sal fin. bluefin tuna, Thunnus thynnus g The most anterior of the chro- 2 matophores on the dorsal side is located anterior to the origin of the second dorsal fin. ...... long-tailed tuna, Thunnus tonggol f One of several melanophores appear along the ventral side of the body. bigeye tuna, Thunnus obesus c No melanophores appear on the sides of the body. h Chromatophores appear at tip of lower jaw. In the profile of the head, the center of the eye is clearly located higher than the tip of the snout. yellowfin tuna, Thunnus albacares h 2 No chromatophores appear on tip of lower jaw. In the profile of the head, the center of the eye is about on the same level as the tip of the snout or slightly higher. albacore, Thunnus alalunga Briefly, larval Auxis can be distinguished from other larvae of tuna and tunalike fishes by the presence of a dotted line of melanophores along the dorsal, lateral, and ventral center lines of the caudal peduncle (Ueyanagi 1964). The standard pattern is three lines, but sometimes only two — the dorsal and ventral — are seen. In small specimens, however, the line on the ventral side is not confined to the vicinity of the caudal peduncle but extends farther forward. Other character- istics include the appearance of melanophores on the isthmus and just anterior to the anus, late appearance of chromatophores on the forebrain (they do not appear until the fish attains a body length of about 8 mm), and poorly developed chromatophores on the first dorsal fin (they do not appear until the fish becomes about 10 mm long). The outline of the anterior surface of the head gives an impression of roundness and the separation betweeen the first and second dorsal fins becomes quite apparant at lengths of 10 mm or greater. Additional observations on the morphology of larval Auxis showed that the snout is shorter than in other tuna and tunalike larvae of comparable size (Jones 1963). Jones also noted that larval Auxis has 39 myotomes, a large chromatophore at the symphysis of the pectoral girdle, and that fin ray development is later than in other tuna and tunalike larvae. Jones and Kumaran (1964) added that compared with larval Euthynnus affinis, larval Auxis exhibits later develop- ment of the first dorsal fin and has relatively few chro- matophores on the first dorsal fin membrane, and its fin membrane is relatively narrow. The variations in pigmentation among larval Auxis has been used as a basis for separating the larvae into two possible types. Matsumoto (1959), who described the morphology of larval and postlarval Auxis on a worldwide basis, pointed out the possibility that this variation may well be due to the presence of two species of frigate tuna; therefore, he provided two separate de- scriptions calling them Type I and Type II, as shown in Figure 22. Type I and Type II larvae show only minor differ- ences and the most obvious one is the variation in pig- mentation near the caudal peduncle (Matsumoto 1959). Type I, which resembles the form described as A. thazard (Matsumoto 1958), has three equally developed rows of pigment whereas in Type II the middorsal edge of the caudal peduncle usually contains only one or two chromatophores and the midlateral line has none. Subsequently, Type I larvae, which are stouter than Type II, were provisionally identified as A. thazard and Type II, which are relatively elongate, were identified as A. rochei (Jones 1963). Further observations revealed that compared to A. thazard, larval A. rochei has a relatively shallow body, shows later development of the spinous dorsal, and has less intense pigmentation on the caudal peduncle (Jones 1963; Gorbunova 1969). A later study confirmed Jones' observation; Yabe and Ueyanagi (1962) published the results of a study on larval tuna in which they positively identified specimens of A. rochei. A controversy over the identification of larval Auxis erupted in 1963 when Jones (1963) questioned the re- sults obtained by Mito (1961), who described the egg and larval development of Auxis (Fig. 23). Jones pointed out that because A. rochei is the most common frigate mackerel in Japanese waters, the eggs and larval stages described by Mito may be this species; but he also noted that there were striking differences between larvae described by Mito and those described by Matsumoto (1959) and Jones (1961). These differences, he believed, should be explained by Mito. The breakthrough in arti- ficial rearing accomplished by the Japanese should make it possible to compare specimens of artificially reared larvae and those obtaind from net tows and night lighting to ob- tain positive identification. 3.23 Adolescent phase Juvenile Auxis can be distinguished from other tunas by certain morphological features (Schaefer and Marr 1948; Wade 1949; Mead 1951; Matsumoto 1962). They can be distinguished from other tunas by the wide gap between the first and second dorsal fins and the virtual- ly colorless appearance of the first dorsal. Compared to Katsuwonus, however, the first dorsals are somewhat similar, both having a lightly pigmented or colorless fin. 27 I MM. I MM. Figure 22.— Auxis larvae Type I, 7.05 mm (A) and Type II, 7.2 mm (B) (Matsumoto 1959). Katsuwonus, nevertheless, possess more pigment in the first dorsal, which is also more or less brown or tan dis- tally, whereas in Auxis, a few scattered chromatophores are found along several of the anterior rays in the distal portion of the first dorsal. Compared with Euthynnus, Auxis are rounder in cross section, relatively less com- pressed laterally, and the length of the head and the caudal region shorter in comparison to total length. Juvenile Auxis also possess a black spot at the isthmus which is seen only in one other genus, Euthynnus. In young Euthynnus, however, the first and second dorsals are continuous. The large interspace between the dor- sals, the low ray count of the first dorsal, and the absence of an elaborate "trellis" of the vertebrae also help to separate Auxis from Euthynnus and Kat- suwonus. The pattern of pigmentation is also a good identify- ing character. Schaefer and Marr (1948) wrote "In the smaller specimens the prominent areas of pigmentation are on the upper and lower jaws, above the snout, around the postero-ventral margin of the orbit, on the upper operculum, between the orbits, along the mid-line of the body, along the bases of the dorsal and anal fins including the finlets, and around the posterior end of the urostyle. Large chromatophores in the peritoneum show through the body wall along the upper half of the body cavity. None of the fins or finlets bears pigment spots, with the exception of the first dorsal. The first dorsal bears a few scattered chromatophores, largely distrib- uted along the spines." Several features of the axial skeleton also are useful in separating the juveniles of partially digested tunas. Pot- thoff and Richards (1970), who studied regurgitated food of terns and noddies from the Dry Tortugas, Fla., observed that damaged or partly digested Auxis and those <30 mm could easily be separated from Euthyn- nus if judged on fin position and gill raker counts. Pot- thoff and Richards revealed that extreme care must be used to examine Auxis specimens because the spines of the first dorsal in some juveniles are subcutaneous and come to the surface in the fin interspace. Furthermore, they noted that the number of gill rakers is evenly dis- tributed over the range for Auxis whereas they are more normally distributed in other species. Counts were as follows: 19 in 38% of the ceratobranchials, 20 in 25%, 21 in 25%, and 22 in 12%. Potthoff and Richards suggested that this kind of distribution may be attributable to the presence of two types of Auxis and their intermediates. 28 Figure 23. — Various stages of development of egg and larvae of Auxis (Mito 1961); 1) Pelagic egg, 7 1 2 h after collecting, 1.04 mm in diameter, oil lobule 0.26 mm; 2) 29-myotome stage, 9'2 h after; 3) 104 h after; 4) 14'2 hr after, shortly before hatching (24°-27°C); 5) empty egg capsule; 6) larva just hatched, 2.70 mm TL, myotomes 11 + 32 = 43; 7) larva 16'/ 2 h after hatching, 3.68 mm TL, myotomes 9 + 31 = 40; 8) larva 30 h after, 3.58 mm TL, myotomes 9 + 31 = 40; 9) larva 42V 2 h after, 3.92 mm TL, myotomes 9 + 32 = 41; 10) larva 55 h after, 3.75 mm TL, myotomes 8 + 31 = 39. in addition to the gill raker counts (38-43 for A. thazard and 43-49 for A. rochei), the form of the free parapophyses is a useful feature of the axial skeleton for separating Auxis to species. Watanabe (1964) deter- mined that in A. thazard, the free parapophyses are long, extend downward and almost touch the haemal spine of the preceding vertebra whereas in A. rochei they are short. This difference in length of the parapophyses is particularly evident on specimens of 111 mm or more (Figs. 24, 25). Four other characters used by Watanabe (1964) in- clude: 1) the number of the vertebra on which the first free parapophysis occurs, 2) the number of the vertebra on which the first inferior foramen occurs, 3) the ratio of the length of the caudal to the precaudal vertebra, and 4) the ratio of the body height at the origin of the anal 29 Auxis thazard Auxis rochei Specimen No. 7 Standard length 88.5 ram. Specimen No. 28 Standard length 94.0 mm. (free parapophysis) Specimen No. 37 Standard length 123.6 mm. Specimen No. 23 Standard length 124.6 mm. Figure 24. — Caudal vertebrae of Auxis thazard (left) and A. rochei (Watanabe 1964). fin to the distance between the origin of the first and sec- ond dorsal fins. Watanabe found that in A. thazard, the first free parapophysis occurred most frequently on the I I Auxis rochei fl Auxis thazard nn n nQ n n [RH nn n nn,n a n Ea ; n ji n n n n n nJ n nn 28 th 27th 26th 25 th 24 th 23 th 11 81 91 101 III 121 131 141 I I I I I I I 90 100 NO 120 130 140 150 STANDARD LENGTH (mm) 23d vertebra, whereas in A. rochei, it was on the 21st or 22d vertebra on 80% of the specimens; however, there were some in which it occurred on the 23d vertebra. Furthermore, the inferior foramen in young specimens occurs on the 27th or 28th vertebra in A. thazard and usually on the 29th and 30th vertebra in A. rochei. Watanabe also observed that the ratio of the length of the caudal to the precaudal vertebra for young speci- mens was 0.973 or less in A. rochei and 0.985 or more for A. thazard. This character also changes as growth oc- curs; therefore, the stage of growth needs to be consid- ered in making a diagnosis. Figure 26 shows a plot of the ratios — body height at the origin of the anal fin to the distance between the origins of the first and second dorsal fins — on standard lengths and except for a single specimen which showed an intermediate value, those for A. thazard tend to be higher than for A. rochei. O.f 0.6 I 1 1 1 ^~ II a a 0.5 0.4 a 75 mm, fishes formed 39% and Crustacea 42% by volume. Squids occurred very infre- quently. Anchoviella sp. and Leiognathus sp. were the AUXIS THAZARD 49-75mm 'Kibinago — Spratelloides japonicus (Houttuyn). -'Ginkagami — Mene maculata (Bloch). AUXIS THAZARD 76- 132mm 1 FISH 1 CRUSTACEA [ : | CEPHALOPODA I MISCELLANEOUS AUXIS ROCHEI 170-252 mm Figure 30. — Diagrams illustrating the composition, by volume, of the stomach contents of Auxis thazard and A. Rochei caught in the Indian Ocean (Thomas and Kumaran 1963). Table 21.— The stomach contents of Auxis rochei caught by trolling gear and set net in Japanese waters. 1958-61 (Yokota et al. 1961). Fishing Num- Spotted Jack Nezumi Kurotachi Sagi- Lizard- Hata- Matouishi- Mishima- Date ground ber mackerel mackerel Saury gisu 1 kamasu fue 3 fish aji 4 mochf okoze" Squid Anchovj Trolling line Dec. 1958 Kumanonada 187 2 2 1 4 Jan. 1959 Kumanonada 137 1 1 36 Feb. 1959 Kumanonada 175 1 2 38 Mar. 1959 Kumanonada 105 4 1 2 1 1 Dec. 1960 Kumanonada 30 243 Jan. 1961 Kumanonada 9 1 Mar. 1961 Kumanonada 29 4 2 Apr. 1961 Kumanonada 17 6 3 25 Set or fixed net: Apr. 1961 Tanegashima 30 37 995 26 459 Nezumi gisu - Gonorhynchus abbreviates Temminck and Schlegel. Kurotachi kamasu - Acinacea notha Bory and St. Vincent. Sagifue - Macrorhamphosus scolopax (Linne). Hataaji - Elephenor macropus (Bellotti). 'Matouishimochi - Apogonichthys carinatus (Cuvier and Valenciennes). Mishimaokoze - Gnathagnus elongatus (Temminck and Schlegel). 35 most common among the fish species consumed. Among 31 preadult A. rochei (170-252 mm), fishes constituted 42%> by volume and were found in 80% of the samples (Fig. 31). Those that were important were Sardinella spp., Anchoviella sp., Leiognathus sp., and carangids (Table 22). Crustaceans were next in importance ac- counting for 24% by volume and found in 77% of the samples. Rabindra Nath (1962) reported that among the most common crustaceans consumed by Auxis in Indian waters were Rhopolophthalmus sp., Hyperia bengalensis, Oxycephalus clausi, Pseudophausia lati- frons, Acetes erythreus, and Squilla larvae. Cephalopods formed 22% of the food consumed (Kumaran 1964), but pteropods were relatively unim- portant in the diet (Rabindra Nath 1962). Kumaran (1964) also noted that larval stomatopods and Lucifer constituted a major portion of the diet of some specimens captured near Quilandy on the west coast of India. Other items of food occasionally seen were chaetognaths, Halobates, and polychaetes. Table 22. — List of food items of preadult specimens of Auxis rochei from the Indian Ocean (Kumaran 1964). Figure 31. — Percentages, by volume, of the types of food con- sumed by preadult specimens of Auxis rochei in the Indian Ocean (Kumaran 1964). 3.43 Growth rate Observations on postlarval growth indicate that cap- tive Auxis grow faster than Euthynnus lineatus (Clemens 1956). Of five postlarval Auxis placed in an aquarium for observation, three died several hours later from injuries received in handling. Of the two remaining Auxis, one measuring 20 mm TL grew to 40 mm in 6 days; the larger 30 mm specimen reached 46 mm during the same period. The rate of growth of A. rochei estimated from modal progression of length-frequency data appears to be rather slow compared to that observed in larvae that were hatched from artificially fertilized eggs. Hotta Number of Percentage food of Percentage Food items organisms prevalence by volume Polvchaeta 14 3.2 1.2 Crustacea: (142) (77.4) (24.4) Amphipods 8 16.1 0.9 Mysis stage of prawn 8 12.9 0.9 Megalopa larvae 118 38.7 21.0 Alima larvae 5 12.9 0.9 Unidentified crustaceans 3 6.4 0.6 Insecta: Halobates 2 3.2 0.6 Chaetognatha: Sagitta spp. 580 3.2 8.7 Cephalopoda: Sepinteuthis sp. 7 16.1 22.7 Vertebrata (Pisces): (78) (80.6) (42.3) Sardinella spp. 5 9.7 3.5 Clupeid larvae 8 6.4 2.9 Anchoviella commersonii 3 6.4 2.3 Anchoviella tri 7 12.9 4.7 Hemirhamphus sp. 3 6.4 2.7 Sphyraena sp. 2 3.2 2.3 Caranx sp. 3 9.7 1.8 Carangid larvae 22 9.7 2.1 Leiognathus sp. 8 16.1 5.3 Unidentified fish and larvae 17 25.5 14.5 (1955), who used monthly length-frequency histograms to estimate the growth of A. rochei caught in the north- eastern sea off the Pacific coast of Japan, constructed a growth curve depicting body lengths at ages 0-IV. From the curve, shown in Figure 32, it appears that A. rochei reach about 17 cm 1 yr after hatching. But Harada, Murata, and Furutani (1973) observed that larvae of A. rochei under the best condition grew to 15.7 cm in 52 days (Fig. 21). In addition to growth rates, the condition factor has been calculated for A. thazard. The condition factor (K), which expresses the relative well-being of the fish, 50 40 JO * ^ X J*J ^^x o o/ / o SATSUNAN • SHIKOKU » IZU ARCHIPELAGO x T0H0KU /. o i 5 i rs AGE (YEARS) Figure 32. — Growth curve of Auxis rochei based on specimens from four localities in Japan (Hotta 1955). 36 increases with age because older fish tend to gain pro- portionately more in weight than in length. The formula for K is K = aW V in which a = constant W = weight in grams L = standard length in millimeters (Rounse- fell and Everhart 1953). At La Linea, Spain, the average values of the condi- tion factor for each length group of A. rochei were usual- ly larger for males than for females (Rodriguez-Roda 1966). Furthermore, the values of K decreased from May to August, but increased in September which coincided with the spawning period. Comparison of K among the size groups showed no significant differences. Ishida (1971), who also calculated condition factor of Auxis caught off Japan, found that of the two species, A. thazard caught at Mikomoto and Shionomisaki had higher indices than A. rochei. Between areas, however, there appeared to be no appreciable difference. The length-weight relationship has been described for A. thazard and A. rochei caught in the world's oceans. The expression for the relationship is W = aL» in which W = weight in grams L = length in centimeters a and b = constants. Table 23 gives the constant for the expression of the predictive length-weight relationship calculated by Ishida (1971) for both species of Auxis and by Yasui (1975) for A. rochei caught in Japanese waters, by Lenarz (1974) for A. rochei captured in the Atlantic Ocean, by Rodriguez-Roda (1966) for A. rochei caught in Spanish waters near the Strait of Gibraltar, and by Sivasubramaniam (1966) for both species caught around Sri Lanka in the Indian Ocean. Figure 33 depicts the length-weight curves constructed by Rodriguez-Roda (1966) and Yasui (1975). For A. rochei caught in Spanish waters, Rodriguez-Roda has also provided the average and calculated weights for the length groups studied (Table 24). The maximum size reported is 66 cm for A. rochei in the eastern Atlantic Ocean (Collignon see footnote 2). Table 24. — Length groups, average weights, and calculated average weights of Auxis rochei caught at Barbate, Tarifa, and La Linea, Spain, in 1958, 1961, 1963, and 1974 combined (Rodriguez-Roda 1966). Size Average Calculated intervals weight average weight (cm) N (kg) (kg) 34-34.5 2 0.635 0.654 35-35.5 10 0.709 0.715 36-36.5 26 0.782 0.780 37-37.5 58 0.864 0.849 38-38.5 97 0.947 0.922 39-39.5 115 1.010 0.999 40-40.5 117 1.089 1.080 41-41.5 88 1.169 1.166 42-42.5 89 1.250 1.256 43-43.5 97 1.339 1.351 44-44.5 34 1.428 1.450 45-45.5 11 1.551 1.555 Total 744 3.44 Metabolism In A. rochei, a correlation has been found between sexual maturity and the amount of iron, copper, and zinc in various body tissues (Suzuki and Morio 1957). Suzuki and Morio showed that in the liver, the iron con- tent decreases progressively with maturation, is at a minimum when the gonads become ripe and then in- creases significantly when the gonads are spent (Fig. 34a). The contents of copper and zinc show a similar tendency to decrease with maturation but minimums are reached at a stage when egg formation is still con- tinuing in the ovaries (Fig. 34b, c). Suzuki and Morio also observed a negative correlation between the fat con- tent of the liver and the amount of iron and copper (Fig. 34d). The relationship, however, is not consistent; it tends to break down in fish that have gonads approach- ing the ripe stage. Concerning molybdenum and nickel in the tissues of A. rochei, Morio and Suzuki (1959) determined that the former occurs in highest concen- tration in the liver whereas the latter shows up highest in the pyloric caeca. Further analysis indicated that the Table 23. — Constants calculated by various investigators for the expression of the predictive length-weight relationship of Auxis thazard and A. rochei. Investigators Auxis thazard Auxis rochei Area N a b N a b Japan: Yasui (1975) — — — NA 1 1.549 X 10 : < 3.65705 Mikomoto Ishida (1971) NA' 6.05 X 10 :1 3.300 NA' 4.13 X 10' 3 3.384 Shionomisaki Ishida (1971) NA 1 7.70 X 10~ 2 2.509 NA' 4.64 X 10" 3 3.362 Atlantic Lenarz (1974) — — — 50 2.80 X 10' 4.13514 Spain Rodriguez-Roda (1966) — — — 744 1.00538 X 10 s 3.129871 Sri Lanka Sivasubramaniam (1966) 160 1.780 X10< 3.3338 28 2.598 X 10" 6 4.6315 Not available in the paper. 37 800 600 o 400 o CD 300 100 (A) W - 1.549 X 10 L 2000 600 h i i — i — i — r (B) W= 1.00538 X icr 5 L 3129871 20 25 30 FORK LENGTH (cm) 35 34 36 38 40 42 44 46 48 FORK LENGTH (cm) Figure 33.- - Length-weight relationship of Auxis rochei caught in (A) Japanese waters (Yasui 1975) and in (B) Mediterranean waters (Rod- riguez-Roda 1966). contents of molybdenum and nickel in the liver show no relationship to sexual maturity, rate of growth, and area of catch. The blood of Pacific A. rochei has been tested and found to be relatively high in hemoglobin concentration (Klawe et al. 1963). Barrett and Williams (1965) showed that compared with other scombrids, the mean level found for A. rochei was among the highest, reaching 19.2 g/100 ml and varying from 16.5 to 22.8 g/100 ml. Blood smears of A. thazard caught in Puerto Rican waters showed a blood count of 1 large hemoblast, 1 small hemoblast, 1 large lymphocyte, 30 small lymphocytes, 45 thrombocytes, and 22 granulocytes (neutrophils) (Saunders 1966). Mature erythrocytes from A. thazard blood averaged 10 /x in length and 7.5 n in width. 3.5 Behavior 3.51 Migrations and local movements Not much is known about the migration of Auxis in the world's oceans. Most of what is known comes from results of studies conducted in Japanese waters. Hotta (1955), after examining the fluctuations in landings of A. rochei along the coast of Japan in 1952, determined their seasonal movement up and down the coast of Japan. He observed that they occur in the Satsunan Sea region and off Shikoku early in the year, but as the year progresses, larger catches are made farther north and by June they are landed in the Izu Archipelago region (Fig. 35). The schools reach their most northern point around Hokkaido in September, then move southward until in December they appear to be concentrated mostly off the Izu Archipelago and Shikoku. Detailed studies of long-distance movement of A. rochei in Japanese waters showed that fish caught, tag- ged, and released in August, September, and November had a general tendency to move southward (Table 25, Fig. 36). Yamashige (1974) determined from his tagging studies that within the population of A. rochei off Japan, differential migration occurs according to size. The large influx of small fish into the fishery, as shown by the length-frequency distribution (Fig. 37), is the consequence of small fish moving south before the move- ment of large fish begins. Short-distance movement of tagged A. rochei in Japanese waters shows that they tend to move south- ward late in the year although some northward move- ment is also seen, but early in the year there is a definite northward movement (Table 25) (Morita 1972; Hamada, Ishida, Morita, Takezawa, Okabayashi, and Ishii 1973; Hamada et al. 1974). 3.52 Schooling The schooling instinct, according to a number of sources, is very strong and orderly in Auxis. Jones (1963) noted that the tendency of A. rochei to form large 38 (o) 1 ' 1 1 X - LIVER CSSSlE *x JL i-' ** X 500 400 300 200 100 GONAD 1 1 X X • \ X •X •X IMMATURE EARLY LATE MATURITY— RIPENING GONAD IMMATURE EARLY LATE MATURITY— RIPENING 5O0 400 300 200 (c LIVER 1 1 1 • • „ • X -' >> X «* • < ^~""'~ """"x • \ X V x --I X X \ IMMATURE EARLY LATE MATURITY— RIPENING RIPE (d 1 • i 1 cc 300 • He p x Cu fef 5 \ • i T XII N CO \ lil \ or \ o 200 \ \ \ \.,3 \ - m . \ 1 \ \ .13 5 x \ \ \ CO \ Jo, ^ 100 \ \l X 1 \ o • 5 • 6 \ »6 \ 3 x5 ^-om — innoo <3.75 3 75 II 25 II 25 37 51 4 _ 10,000 — 35,000 5_35,000 -105,000 6 105 000-350 000 37.51 131.29 131.29—393.88 393.88- 1,312 94 7 > 350 non ^ 1 31? 95 JUNE ^ i& SEPT. x fte- 1 OCT. jj ^ ft J J SL (4 at?- » jpx JULY Figure 35.— Landings of Auxis rochei off Japan, by month, in 1952 (Hotta 1955). NOV. £* n r 7 $ Figure 36. — Long distance movement of Auxis rochei tagged in August, September, and November (Yamashige 1974). HONSHU • RELEASE * RECAPTURE -34°- 32" catches showed a slight predominance of females during periods of southwest monsoons; the male to female ratio reached 1:1.5 during this period. For A. rochei caught in Spanish waters, Rodriguez- Roda (1966) also tested the ratios of males to females in the catches at Barbate, Tarifa, and La Linea, particu- larly for those months where a ratio of 1:1 was in doubt. The results showed no significant departures from a 1:1 ratio except at La Linea where in September 1961, sig- nificantly more males were taken in the traps. 42 30 20 10 30 20 10 30 20 10 30 20 10 30 20 £ 10 § 40 " 30 "- 20 g '0 f* g 30 O 20 20 10 30 20 10 30 20 10 30 20 10 20 10 JAN. FEB. ^T-TmH~T^ MAR. APR. MAY r^-J j n^^ NOV. JHH DEC. 18 20 22 24 26 28 30 32 34 36 FORK LENGTH (cm) Figure 37.— Percentage length-frequency distributions, by month, of Auxis spp. taken in Japanese waters (Yamashige 1974). 4.13 Size composition Although the size of A. rochei in the Japanese com- mercial catch varies widely from 4 to 50 cm, the bulk of the catch usually falls between 20 and 35 cm (Fig. 38) AUGUST(N=559) otik SEPTEMBER (N* 824) 0CT0BER(N = I94) NOVEMBER (N-= 138) DECEMBER (N = 112) I I JANUARY (N= 32) FEBRUARY (N = I6) MARCH (N= 89) APRIL(N=47) MAY(N = 52) JUNE (N-=36) JULY(N = 22) ■ TOHOKL r-| K^| H SATSUN I I SHIKOK H L ■_ n 30 gross tons which fish in the coastal waters and high seas for surface schools of tunas such as skipjack tuna, yellowfin tuna, and alba- core. Japanese pole-and-line boats that fish in the At- lantic for yellowfin and skipjack tunas and occasionally for Auxis and bigeye tuna vary from 151 to 239 gross tons, whereas the seiners, both single and double boats, range from 50 up to >400 gross tons (Borgstrom 1964; Hayasi 1973, 1974). Among the more primitive types of crafts used in Auxis fishing are dugout canoes and catamarans (Silas 1967b). In the Maldive Islands, for example, sailing boats called vadu dhony, which are about 6.1 m long and 1.8 m wide, and fish three to four baited trolling lines, are used as trollers (Sivasubramaniam 10 ). Slightly larger sailing crafts, called the mas dhony, are used in pole-and-line fishing; they vary in length from 10.7 to 12.2 m, have a beam of about 3.4 m and a draft of nearly 0.8 m. The Maldivian boats are constructed of coconut wood, beautifully streamlined, and keeled for wind- ward sailing. Compartments for carrying live bait have continuous water circulation through holes along the bottom. The fishermen fish at the stern atop a U-shaped removable wooden platform. Figure 50. — Percentage composition of tuna species caught by various types of gear from the coastal waters of Sri Lanka (Siva- subramaniam 1965). 10 K. Sivasubramaniam, UNDP-Sri Lanka Skipjack Fishery Develop- ment Project, FAO, Colombo, Sri Lanka, pers. commun. August 1975. 50 In 1972, the Maldivian fishing fleet had an estimated 5,100 fishing boats which included 2,980 trailers and 2,100 pole-and-line tuna boats operating within a 46 km radius of the islands. Skipjack tuna, yellowfin tuna, kawakawa, and Auxis are the main species caught by the pole-and-line boats. Hiebert and Alverson (1971) pointed out that one of the factors hindering the development of the Maldivian tuna fishery is the limited access of the boats to more distant fishing grounds. Lack of refrigeration also precludes long trips; therefore, the tuna boats leave port for the fishing grounds at about 0300 and return late in the evening. The introduction of motorized, refrigerated boats can probably increase the present fish catch 3-4 times. hi waters around Sri Lanka, trolling gear was formerly operated from "orus" or outrigger canoes with sails. Nowadays, however, 3- to 5-gross ton motorized boats, manned by a crew of at least three fishermen, are used (Sivasubramaniam 1973). In the Canary Islands where Auxis are landed along with other species of tuna, there are large boats (average 400 gross tons), which have re- frigerated holds and carry about 20 men, and smaller, artisanal fishing boats (average 10 gross tons), which are crewed by 4-5 fishermen, are about 12 m long and have open decks and bait tanks. Propelled by engines that vary in size from 50 to 120 hp, these small boats usually engage in pole-and-line and occasionally purse seine fishing. Off Angola, yellowfin and skipjack tunas are the principal species caught by the small, pole-and-line boats, but Auxis are also landed in smaller quantities (De Campos Rosado 1972). These boats, which operate within a narrow strip about 75 km wide and 370 km long along the coast, vary in size from 12 to 20 m long. The Cuban pole-and-line fishery, which lands small amounts of Auxis, operates motorized sailing boats, which have been described as modified sloops with a gaff-rigged mainsail and usually with a flying jib (Rawl- ings 1953; Wise and Jones 1971). 5.2 Fishing areas 5.21 General geographic distribution Almost all the Auxis caught in the Japan Sea are A. rochei (Okachi 1958). Percentagewise, Okachi deter- mined that of the fishing areas established for Auxis in Japanese waters, the South Pacific region had 62% of the catch, the middle Pacific region 22%, and the Japan Sea and East China Sea regions combined only 12%. Catches from west Japan Sea and East China Sea regions were usually similar whereas catches from the north Japan Sea region were smaller than either one. In other areas of the world, Auxis are usually caught as incidental species. In the Pacific, for example, Auxis are caught incidentally in waters around Taiwan, the Philippines, Thailand, West Malaysia, Hawaii, and Australia (Serventy 1941; Gosline and Brock 1960; Kume 1973; Philippine Bureau of Fisheries 1973). Small numbers of Auxis are also taken in the Indian Ocean. Jones (1967) reported that Auxis are captured along the coasts of India and in waters surrounding the Maldive Islands (Sivasubramaniam footnote 10). According to Sivasubramaniam (1973), Auxis are caught in waters around Sri Lanka on a commercial scale throughout the year in the south and southwest coast of Sri Lanka. In the Mediterranean region, the waters around Cyprus, Greece, Italy, Malta, Morocco, Spain, and Yugoslavia are fishing grounds for Auxis. Spanish trap- net fishermen land commercial quantities of Auxis at Barbate on the Atlantic Spanish coast near the Strait of Gibralter, at Tarifa in the middle of the Strait, and at La Linea on the Mediterranean Spanish coast near the entrance of the Strait. Some fishing is also carried on around the Canary Islands (Bas 1967; Cendrero and Garcia-Cabrera 1972). Auxis are taken in Moroccan waters mostly by trap fishermen. Traps located at Larache and at Cape Spartel are set primarily to catch bluefin tuna that make their spawning migration toward the Mediterra- nean Sea along the Moroccan coast in April- July. Traps at M'diq in the Mediterranean also capture tuna as they exit after spawning. In traps located along the Atlantic Moroccan coast, Auxis account for only 3% of the tuna landed, but in those situated along the Mediterranean Moroccan coast, they represent 98% of the tuna catch (Lamboeuf 1972). Around Portugal, the tuna fishing grounds are along the north and west coasts where small pole-and-line ves- sels operate. Other grounds fished by Portuguese fisher- men are located off the islands of Azores and Madeira where small- to medium-sized pole-and-line vessels operate, and off Angola (De Campos Rosado 1972; Dias and Barraca 1972). Catches of little tunny, bonito, and Auxis represent roughly 60% of the total tuna landings from Angola. Off Ghana, in the Gulf of Guinea, there is another fishing ground for Auxis. Although most of the tuna landings at Tema are made by foreign longline, pole- and-line, and purse seine vessels, Ghana's canoes and motorized fishing crafts also account for some of the landings of tuna and tunalike fishes (Di Palma 1968). In the western Atlantic, Auxis were previously taken along the east coast of the United States. Pound nets fished along the Middle Atlantic States during the 1940's and early 1950's frequently caught Auxis along with bluefin tuna, little tunny, and bonito (Anderson et al. 1953). Southward, the waters around Cuba, off Venezuela, and off Brazil also yield small quantities of Auxis. 5.22 Geographic ranges See section 5.21. 5.23 Depth ranges Adult A. thazard has been reported to distribute themselves vertically from the surface down to a depth of about 45 m (KisKinouye 1923). 51 5.3 Fishing seasons 5.31 General pattern of seasons The average monthly landings of Auxis at Tosashimizu (Kochi Prefecture), at Hachijo (Izu Archipelago), and at Mera (Shizuoka Prefecture) are shown in Figure 51. At Tosashimizu, where a pole-and- line fishery operates, Auxis are caught throughout the year, but peak fishing usually extends from October through May (Yasui 1975). During the remainder of the year (June-September), some of the pole-and-line boats concentrate on other species; therefore, there is a reduc- tion in fishing effort for Auxis. In the region of the South China Sea, the fishing season for Auxis varies considerably. For example, in the Philippines, the fishing season varies for different parts of the island group beginning at any time between November and January and extending into May (Philippine Bureau of Fisheries 1973). In Thailand and West Malaysia, however, the landings of Auxis, Euthyn- nus affinis, and Thunnus tonggol are usually better dur- ing the latter half of the year (Kume 1973). Fishing for A. thazard in Indian waters begins about August and extends to about December (Nair et al. 1970). Sivasubramaniam (1973), who examined the 600 HACHIJO i— | ! 1 l i i i 400 " 200 - | 1 | , - ru.._ ,n i i i i i i MERA i i i i 1 1 1 1 i ! ' ' 1 1 1 1 W 500 < oz JAN. FEB MAR APR MAY JUNE JULY AUG SEPT OCT NOV DEC. Figure 51. — Landings of Auxis spp. average by month from catches made between January 1969 to December 197G at Tosashi- mizu, between January 1969 and October 1974 at Hachijo, and between January 1967 and December 1974 at Mera, Japan (Yasui 1975). seasonal and annual variations in the catches of frigate tuna from waters surrounding Sri Lanka, observed that in the south and southwest coasts, A. thazard are caught throughout the year but in the east and north coasts, they are caught primarily in June-September and lesser amounts in March-April. In the Maldives, catches of Auxis, combined with that of kawakawa, peak at least twice a year (Hiebert and Alverson 1971). In 1965, for example, combined catches of Auxis and kawakawa in the Maldives peaked in January at 1.2 million fish, declined sharply in February-May to <0.6 million, then fluctuated ir- regularly until September after which the catch rose to a second minor peak in November before declining again in December. At Barbate, Spain, A. rochei appear in the trap catches in May-July with the heaviest catches occurring in the first 2 months. About 53% of the catch is com- posed of A. rochei and the remainder consists of little tunny and bonito (Rodriguez-Roda 1966). At Tarifa, bullet tuna constitute 95% of the trap catches and average 67.1 t annually. Here, the trap fishery is active in May- June. The traps at La Linea make the largest catches of small tunas with A. rochei accounting for 97% of the catches. The fishing season extends from August to October but most of the catches are made in the first 2 months. 5.32 Dates of beginning, peak, and end of season See section 5.31. 5.33 Variation in date or duration of season At Tosashimizu, Japan, where a pole-and-line fish- ery for A. rochei operates, severe fluctuations in the monthly landings usually occur in October-December whereas landings are stable in January-March (Yasui 1975). Yasui examined the relationship between sea- surface temperature and monthly landings and found that in years of good fishing, there was a complex distri- bution of surface temperature with 28°C water very close to shore (Fig. 52A), whereas in years of poor fish- ing the warm water was displaced farther offshore (Fig. 52B). He concluded that the location of the 28° C isotherm in July-September significantly influenced fishing conditions in subsequent months. Temperature, it appeared, affects the concentration of bullet tuna schools, i.e., their availability to the boats and the length of their stay on the fishing grounds. Monthly landings at Mera, where the fish are taken by nets, show that peak fishing occurs in July-Septem- ber (Fig. 51). Yasui (1975) noted that in this fishery, the large, mature fish with high reproductive index are usually taken in June and that progressively smaller fish are taken in subsequent months. The proportion of im- mature fish in the catch increased in July-August and by September all fish were immature. Yasui concluded from these findings that the catch of Auxis at Mera is in- 52 Figure 52. — Distribution of surface temperatures in July- September in years of good (A) and poor (B) fishing for Auxis rochei at Tosashimizu, Kochi Prefecture, Japan (Yasui 1975). fluenced to a large extent by a seasonal in-migration of fish from the west and south. At Hachijo, where Auxis are landed by surface trail- ers, catches are made throughout the year except in Oc- tober (Fig. 51). Yasui (1975) observed that there are two seasons for Auxis at Hachijo — one in April-August and another in November-January. The appearance of Auxis in April-May at Hachijo coincides with the end of the gill net fishery for flyingfish, which are preyed on by Auxis. Yasui also noted that sexually mature fish appear in the catches in April-May followed by larger, older fish in July. At Ashizurimisaki in Kochi Prefecture, Ishida (1972b) noted considerable fluctuations in landings of Auxis. To determine the possibility of predicting the degree of suc- cess of a fishing season, Ishida, using data for 1952-67., obtained the following: Cj = catch in October-March in year; C J+1 = catch in October-March in year) + 1 AC J+ 1 = difference between C t and C J+ 1. Plotting Cj against AC J+1 , Ishida found that when C } was small, AC J+1 tended to be positive and vice versa. Ishida also noted that larger catches of Auxis tended to be associated with cooler temperatures. Yamashige (1974), on the other hand, found a positive relatonship between catches of A. rochei in August- July and the dif- ference in the mean surface temperature between August and October of a given year, and he described the relationship between these two variables as follows: Y = 1.42 + 1.15 (t 8 - t 10 ) where Y = catch, in 1,000 t, from August (n year) to July (n + 1 year) t s = mean surface temperature in August at Ashizurimisaki (°C). t lQ = mean surface temperature in October at Ashizurimisaki (°C). In Sri Lanka, the mechanization of fishing boats made fishing possible even during periods of monsoons and has been primarily responsible for a shift in the fish- ing season. Originally, Williams (1963) reported that the main fishing season for Auxis in Sri Lanka extended from October or November to May. Sivasubramaniam (1973), however, found that Williams defined the fish- ing season by examining catches made during the north- east and intermonsoons when beach seines are usually in operation. Excellent fishing occurs in all fishing areas around Sri Lanka particularly during the end of the southwest monsoon. Sivasubramaniam observed that this period is associated with heavy recruitment and vulnerability of the recruits to trolling gear. The result is that the number of fish caught per day rises abruptly, but the net tonnage landed, does not rise proportion- ately. 5.4 Fishing operation and results 5.41 Effort and intensity In the Auxis fishery off Ashizurimisaki in Kochi Pre- fecture, Japan, fishing intensity (the number of small, 3-gross ton pole-and-line boats operating per day) varies widely from as few as 10 to as many as 260 boats/day (Ishida 1972a). Fishing intensity usually peaks once in March and again in November to about 200 boats/day. During the balance of the fishing season, which extends usually from October to May, however, the average number operating is about 170 boats/day. 5.42 Selectivity Although pole-and-line fishing is effective for skip- jack and yellowfin tunas, it is not equally effective for Auxis. Sivasubramaniam (1973) observed that in Sri Lanka, Auxis respond poorly to chumming probably because the live bait used is larger than the most com- mon food items in their stomachs. The drift net, which is in operation around the entire island of Sri Lanka, is also ineffective at times par- ticularly for Auxis <30 cm. It is, however, effective for skipjack and large kawakawa. Catches of Auxis by drift nets, therefore, amount to only about 18 kg/day for a 3.5- 53 gross ton boat and about 54 kg/day for an 11 -gross ton boat. Auxis rochei, being smaller, are very rare in drift net catches. 5.43 Catches Total catches of frigate and bullet tunas by various countries in the Atlantic (including the Mediterranean Sea), Pacific, and Indian Oceans in 1953-76 are given in Table 26. By far, the total catch in the Pacific was the largest among the three oceans, averaging 40,700 t an- nually. The Atlantic and Indian Ocean catches, on the other hand, averaged about 13.8 1 in each ocean or about a third of the Pacific catch. In the Report of the Ad Hoc Committee Meeting of Specialists to Review the Biology and Status of Stocks of Small Tunas, it was pointed out that the catches of Auxis reported by the various countries do not truly re- flect the abundance of the two species in the world's ocean. For example, purse seine fishermen avoid catching Auxis because they are a nuisance as many become gilled in the net and require considerable time and manpower to remove. Furthermore, the catches of Auxis by tuna seiners are often discarded at sea. Other difficulties noted with Auxis catch statistics are that ar- tisanal catches are usually not fully reported and that there is confusion and inclusion of Auxis with other species. Pacifc Ocean Japan lands, by far, more Auxis than any other country and is, in fact, the only country that has a well- established commercial fishery. In 1953-76, the catch of Auxis averaged 24,300 1, varying from 14,900 1 in 1975 to 48,300 t in 1963 (Table 26). In Taiwan, the catches of frigate and bullet tunas in 1961-73 varied between 600 and 1,200 t and averaged 1,000 1 (Table 26). Perhaps it should be pointed out that there is a discrepancy in the catch statistics for frigate Table 26.— Estimated catches of Auxis sp. by countries, 1953-76 (in units of 1.000 metric tons). (Data compiled from FAO 1959, 1964, 1969, 1974, 1977; Lamboeuf 1972; Miyake and Tibbo'; Miyake et al. 2 ; Sivasubramanianr'; Hagborg.') — = data not available; 0.0 = magnitude negligible or insignificant. Atlantic Ocean (includir ig Mediterranean Sea) Pacific Ocean Indian Ocean H CO CD >- a o cm 3 00 D u O. >> c CO O u CO 4) o "3 t. o oj'rs c o Is a "3 2 o u o o o 8 '3 a, en 00 co -a '2 "a; M c > 2 > .2 o cm 3 >■ T3 c CO -~ c.S j3 2 o E c CO n. CO CO c 'S.OO Q.T3 — C PL, ,3 a 03 '3 CO a> -a J5 *bi) c C3 CD > to ca o3 S3 c C3 O0 CO Pm c (23 CO c CO 'i-t .2 'c CO c 03 1953 6.4 — — 0.8 — — — 4.0 2.0 — 1.2 — — 15.6 — 1954 7.3 — — 0.6 — — — 1.4 4.0 — 1.3 0.1 — 20.6 — — — — — — — — — 1955 5.2 — — 1.2 — — — 4.0 4.8 — 1.3 0.2 — 23.4 — — — — — — — — — 1956 1.8 — — 0.9 — — — 2.4 3.4 — 1.1 0.1 — 25.9 — — — — — — — — — 1957 1.1 — — 0.5 — — — 2.5 5.4 — 1.0 — — 20.4 — — — — — — — — — 1958 2.5 — — 0.7 0.4 — — 3.0 10.0 — 1.2 0.1 — 23.3 — — — — — — — — — 1959 1.9 — — 0.7 0.7 — — 1.0 3.8 — 1.7 0.1 — 19.9 — — — — — — — — — 1960 1.9 — — — 0.6 — — 1.8 5.2 — 1.3 0.0 — 15.8 — — — — — — — — — 1961 2.7 — — — 1.0 — — 0.2 3.4 — 0.8 0.0 — 18.2 — 0.7 — — — — — — — 1962 1.6 — — — 0.7 — — 0.6 4.4 — 1.0 0.0 — 21.1 — 0.8 — — — — — — — 1963 1.3 — — — 0.8 — — 1.6 3.1 — 1.0 0.0 — 48.3 — 0.6 — — — — — — — 1964 0.9 — — 0.6 0.5 — — 1.5 2.5 — 1.4 0.0 5.3 26.9 9.1 1.1 0.8 2.4 1.5 1.1 — 4.0 — 1965 1.7 0.0 2.0 0.7 0.7 0.9 0.0 1.8 2.5 — 1.8 0.0 4.8 29.8 12.4 0.8 0.4 2.0 2.5 1.2 — 4.5 — 1966 1.4 0.0 2.0 0.5 0.9 0.4 0.0 0.8 2.3 — 1.4 0.0 4.0 29.3 16.4 1.2 0.3 2.0 3.0 1.5 — 4.5 — 1967 1.2 0.0 2.0 0.6 1.2 0.6 0.0 1.2 3.5 — 1.1 0.0 4.4 28.7 10.5 0.8 0.1 1.7 3.0 1.3 — 5.0 — 1968 0.6 0.0 1.8 0.5 1.2 1.6 0.0 0.9 1.7 — 0.4 0.0 3.2 21.0 19.5 0.4 0.1 2.3 3.0 1.3 — 5.0 — 1969 0.8 0.0 3.0 — 1.1 3.2 0.0 0.6 2.0 — 0.4 0.1 3.2 24.1 14.8 0.4 0.2 1.9 3.0 1.0 — 5.5 — 1970 0.5 0.0 3.0 — 1.1 3.1 0.0 1.0 2.2 — 0.7 0.0 2.4 25.9 9.6 0.7 0.7 7.9 1.7 5.5 0.1 5.0 — 1971 1.1 0.0 2.7 — 1.6 — 0.0 0.3 3.8 — 0.5 0.0 2.6 20.1 9.9 0.5 0.7 0.2 1.8 5.6 0.2 3.2 0.4 1972 1.6 0.0 5.3 — 1.7 0.0 0.0 0.5 1.9 — 0.6 0.0 — 31.1 10.6 0.6 0.5 0.2 3.1 4.9 0.2 4.1 0.6 1973 1.1 0.0 2.3 — 1.2 1.2 0.0 1.6 1.9 — 0.7 0.0 — 33.5 13.0 0.7 — 0.2 6.2 5.4 — 3.8 — 1974 1.5 0.0 6.3 — 1.3 0.5 0.0 0.5 0.6 0.0 0.9 0.0 — 26.2 5.0 — — — 5.9 — — — — 1975 0.5 0.0 6.0 — 0.9 0.0 0.0 0.1 0.5 0.0 1.0 0.0 — 14.9 7.6 — — — 3.9 — — — — 1976 0.5 0.0 4.3 — 0.9 0.0 0.0 0.4 0.5 0.0 1.2 0.0 — 20.2 14.0 — — — 2.7 — — — — Mean 2.0 0.0 3.4 0.7 1.0 1.2 0.0 1.4 3.1 0.0 1.0 0.0 3.7 24.3 11.7 1.0 0.4 2.1 3.2 2.9 0.2 4.5 0.5 'Miyake, M. P., and C. G. Tibbo (compilers. 1972. Statistical bulletin. International Commission for the Conservation of Atlantic Tuna ST/Total/72/2-10-11. [No pagination.] 'See text footnote 12. See text footnote 11. D. W. Hagborg, Fishery Resources & Environment Division, Food and Agriculture Organization of the United Nations, Rome, Italy, pers. commun. February 1975. Includes bluefin tuna. 6 Includes Atlantic bonito. 'includes catches from Ceuta in 1954-58. 8 Includes Atlantic bonito in 1965-67. 'Probably includes several other species of tuna and tunalike fishes in 1964-70. 54 and bullet tunas from Taiwan for 1961-68. Examination of the catch statistics from Taiwan for 1958-63 (FAO 1964) showed catches to be rather high. In 1961-63, for example, they were as follows: Year 1961 1962 1963 Catch (1,000 metric tons) 11.9 18.2 15.1 But in FAO (1967), the frigate and bullet tuna catches in 1961-63 were given as follows: Year 1961 1962 1963 Catch (1,000 metric tons) 0.7 0.8 0.6 Further research into the method of compiling catch statistics indicated that data from FAO (1964) included other species under frigate and bullet tunas. Kume (1973) revealed that Taiwan uses a categroy called "bonito" in which are included Auxis as well as kawa- kawa and perhaps other small tunas. Kume gave the "bonito" catch for 1970 as 15,500 1 but estimated that of this total, only 778 1 or roughly 5% were Auxis. Further- more, the frigate and bullet tuna catches in 1961-63, as given in FAO (1967), amounts to only 4-6% of the catches as given in FAO (1964). Therefore, it can be presumed that roughly 5% of the catches of frigate and bullet tunas in Taiwan, as given in FAO (1964), are ac- tually of this species with the remainder constituting catches of other small tunas. Southward, in the South China Sea region, the Philip- pines, Sabah, Sarawak, Thailand, and West Malaysia also harvest Auxis. In the Philippines, catches of Auxis in 1964-76 varied from 5,000 t in 1974 to 19,500 t in 1968 and averaged 11,700 t (Table 26). The Thailand landings of Auxis, Euthynnus affinis, and Thunnus tonggol are combined and called "bonito" or "pla o" locally in the catch statistics (Kume 1973). In 1971, Thai and Chinese seiners landed 5,090 t or 78% of the 6,548 t of "bonito" reportedly caught in Thailand. In West Malaysia, landings of Auxis are also combined with those of E. affinis and T. tonggol. The 1958-71 produc- tion of these three species combined varied between 789 t in 1959 and 5,578 t in 1967 and averaged 3,131 t an- nually. Table 26 shows that the catches of Auxis from China mainland varied from 2,400 to 5,300 t in 1964-71 and averaged 3,700 t. Indian Ocean In addition to E. affinis and T. Tonggol, which are the two most common tuna species landed from Indian waters, small numbers of Auxis are also captured in the tuna fishery along the coast of India (Jones 1967). Of the two species, A. thazard are more common whereas A. rochei are rarely seen in the commercial catch. Thomas (1967) reported that Auxis are taken in the Laccadive Islands, but catches are usually recorded in the category, "other fishes," which includes E. affinis, Acanthocybium solandri, Istiophorus gladius, Elagatis bipinnulatus, Caranx sp., Chorinemus sp., and sharks. Sivasubramaniam" estimated that Auxis catches in In- dia account for about 5% of the tuna and tunalike catches. It is also apparent from the data that the west coast of India usually yields Auxis catches that are about 5 times more than those from the east coast. In 1964-73, India's Auxis catches varied between 200 and 7,900 t and averaged 2,100 t annually (Table 26). The tuna fishery in the Maldives is carried out by sail- ing boats; therefore, fluctuations in the catches are asso- ciated with various environmental conditions such as wind direction and velocity and ocean surface currents. In 1964-76, the catches of Auxis from the Maldives fluc- tuated between 1,500 t in 1964 and 6,200 t in 1973 and averaged 3,200 t annually (Table 26.) There are, how- ever, serious discrepancies in the catch statistics. For 1970 and 1971, for example, Sivasubramaniam (see foot- note 11) gave the catch of Auxis from the Maldives as 1,700 and 1,800 t, respectively. Data from a tuna can- ning plant project, however, indicate catches of 3,094 t in 1970 and 26,871 t in 1971. Based on landings in 1965- 69 and in 1972-73, it appears that the 1971 catch of 26,871 t is unreasonably high. Furthermore, FAO (1974) estimated the 1970 frigate and bullet tuna catch to be 20,000 t and the 1971 catch as 26,900 t, also un- reasonably high; therefore, until these discrepancies can be resolved, data from Sivasubramaniam (see footnote 11) will be accepted provisionally. In waters around Sri Lanka, Auxis are caught on a commercial scale from all the main fishing grounds (Sivasubramaniam 1973). Being the smallest member of the tuna and tunalike fishes, Auxis contribute only 15- 20% to the total catch, by weight, but are the most abundant of all the tuna varieties in the waters around Sri Lanka. Sivasubramaniam reported that although both species appear in the commercial catches, A. thazard contributes 92% and A. rochei 8% to the total annual production. In 1964-73, Sri Lanka's production of frigate and bullet tunas ranged between 3,200 and 5,500 t and averaged 4,500 t annually (Table 26). Per- centagewise, they constituted more than a third of the catch of tuna and tunalike fishes from Sri Lanka in 1964-67. The proportion fell to about a fourth in 1968-69, one fifth in 1970, and has been about one-sixth of the total production since 1971. Other countries that harvest Auxis in the Indian Ocean include Bangladesh, Pakistan, Reunion Island, and Tanzania. Table 26 shows that in 1964-72, Auxis catches in Bangladesh fluctuated between 100 and 800 t and averaged 400 t. In Pakistan, catches in 1964-73 "Sivasubramaniam, K. 1974. More recent information on the tuna fishery in Sri Lanka, India and the Maldive Islands. Indo-Pacific Fish- ery Council/Indian Ocean Fishery Commission Ad Hoc Working Group. Nantes, France, 16-18 September 1974. (IPFC/IOFCAVPU4 12, 4 p. 55 varied from 1,000 to 5,600 1 and averaged 2,900 1 annual- ly. Catches from the Reunion Island, recorded for 1970- 72, are relatively small, varying between 100 and 200 t and an annual average of about 200 t. Tanzania landed 400 t in 1971 and 600 t in 1972 and averaged 500 t for these 2 years. Atlantic Ocean In the Atlantic, Spain, which annually lands about 3,500 t of Auxis, exploits mackerel, Scomber scombrus; bluefin tuna; bonito, Sarda sarda; and occasionally albacore along the Mediterranian Spanish coast (Bas 1967). Mackerel are caught by ring net and trawl; blue- fin tuna, bonito, and Auxis are caught occasionally by very large ring nets. The annual production of Auxis by large ring nets is small, ranging between 274 and 2,500 1. The most important gear is the trap. At Barbate, A. rochei appear in the traps in May-July and catches average 15.6 t but larger catches usually occur in May- June (Rodriguez-Roda 1966). Little tunny and bonito are also taken at Barbate, but A. rochei predominate, accounting for about 53% of the production. The traps at La Linea make the largest catches of these small tunas with A. rochei constituting about 97% of the catches (216 t). At Tarifa, 95% of the catch is A. rochei and production averages about 67.1 t. The annual catches of A. rochei by traps throughout Spain vary widely. Catches at Barbate in some years are very low and fall below 2,000 fish/yr but in other years they may be very high. Table 26 gives the annual production of Auxis in Spain and Table 27 gives a partial breakdown of the catches by type of gear. In 1953-76, the annual landings varied from 500 to 10,000 1 and averaged 3,100 1 (Rodriguez-Roda 1966, 1967). Auxis are taken in Moroccan waters mostly by trap fishermen, but surface gear is also employed (Table 27). Lamboeuf (1972) stated that both bait boats and seiners operate as a team in fishing. The bait boat, which is usually an old, low tonnage sardine boat converted to carry bait, chums fish to the surface and keeps them there long enough for the seiner to set the net around the school. Lacking such a bait boat frequently results in an unsuccessful purse seine set. Along the Atlantic Moroc- can coast, boats landed 69% of the Auxis whereas traps contributed less than half or 31% (Lamboeuf 1973). Along the Mediterranean Moroccan coast, however, traps contributed 91% of the Auxis landed with the remaining 9% produced by boats. In 1953-76, the annual catches of Auxis fluctuated between 100 and 4,000 1 and averaged 1,400 t (Table 26). Auxis landings in Morocco rose from about 15% to 20% of the total tuna landings in 1963-65, fell to about 12% in 1966, then rose steadily to about 27% of the total catch in 1969 (Lamboeuf 1972). The change in the relative importance of Auxis in the total tuna landings resulted from a very sharp decline in bluefin tuna landings in Morocco. It should be noted that slight discrepancies exist in the catch data for Auxis; only 200 1 in Table 26 versus a 1969 catch of 588 1, according to Lamboeuf (1972). In Portugal, traps, which are the principal tuna-fish- ing gear, are usually fished off the southern coast where- as pole-and-line boats operate along the north and west coasts and off the islands of Azores and Madeira (Dias and Barraca 1972). In the Portuguese overseas province of Angola where the yield of tuna is much higher, the fishery is carried on by small, pole-and-line boats (De Table 27.— Annual catches of Auxis (in units of 1,000 metric tons) in the Atlantic Ocean, by country and type of gear, 19G0-74 Type of gear 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 Total 9.0 8.1 8.3 7.9 6.8 9.3 7.1 9.2 9.5 11.0 12.3 10.7 12.8 9.3 (10.7) By type of gear Total surface 4.2 2.8 4.3 4.3 2.6 5.1 3.4 5.8 4.7 5.5 6.2 1.1 1.5 5.9 ( 8.5) Bait boats — — — 1.3 0.9 2.6 1.8 1.8 2.2 4.0 3.1 0.2 0.2 1.2 ( 0.5) Purse seiners — — — — — — — 0.4 1.3 0.2 0.7 0.8 1.2 1.8 — Unspecified 4.2 2.8 4.3 3.0 1.7 2.5 1.6 3.6 1.2 1.3 2.4 0.1 0.1 2.9 ( 8.0) Total traps 1.0 0.8 0.7 1.8 2.3 1.7 1.4 1.1 1.4 0.9 1.1 0.7 1.1 0.3 ( 0.9) Total unclassified 3.8 4.5 3.3 1.8 1.9 2.5 2.3 2.3 3.4 4.6 5.0 8.9 10.2 3.1 ( 1.3) By country Ghana Purse seine — — — — — — — — — — — — — 1.6 — Unclassified — — — — — — — — 1.8 3.0 3.0 2.7 5.3 — 6.3 Japan Purse seine — — — — — — — 0.4 1.3 0.2 0.7 0.7 1.2 0.2 0.0 Bait boat — — — — — 0.9 0.4 0.6 1.6 3.2 3.1 — 0.0 1.2 0.5 Morocco 2 Traps 1.8 02 0.6 1.2 1.4 1.2 0.6 0.6 0.8 0.1 0.5 0.2 0.4 0.3 0.4 Surface gear — — — 0.5 0.1 0.5 0.1 0.6 0.1 0.1 0.5 0.1 0.1 1.3 0.1 Portugal (Angola) Traps — — — — — — — — — — 0.3 0.4 0.4 — — Bait boat 1.9 2.7 1.6 1.3 0.9 1.7 1.4 1.2 0.6 0.8 0.0 0.2 0.2 — — Unclassified (surface) gear — — — — — — — — — — 0.2 0.5 1.0 1.6 1.6 Spain 3 Traps 1.0 0.8 0.7 0.6 0.9 0.5 0.8 0.5 0.6 0.8 0.3 0.1 0.3 — 0.5 Unclassified 4.2 2.6 3.7 2.5 1.6 2.0 1.5 3.0 1.1 1.2 1.9 3.6 1.6 1.9 — Purse seine — — — — — — — — — — — 01 — — — 'Data for 1960-62 from Miyake and Tibbo (see footnote 1 in Table 26); for 1963-73 from Miyake et al. (see text footnote 13); for 1974 from Miyake (personal communication with R. S. Shomura, Southwest Fish. Cent., Natl. Mar. Fish. Serv., Honolulu, HI 96812). z Catch in 1960 includes bonito. 'Catches in 1960 and 1961 include a small amount of little tunny. 56 Campos Rosado 1972). Landings of Auxis from Angola in 1953-76 varied from 500 to 7,300 t and averaged 2,000 t (Table 26). Ghana's modern fishing fleet and industry, which developed only since the early 1960's, are probably the best among African nations (Di Palma 1968). Di Palma observed that confusion exists in the catch statistics on mackerels and scads because some boats do not differen- tiate between the various species. For example, several species may be reported together including mackerel, Scomber japonicus, various species of horse mackerels of the family Carangidae, Auxis, and mackerel scad, Decapterus rhoncus. Miyake et al. 12 reported that Ghana's landings of Auxis by unclassified gear varied between 1,800 and 5,300 t and averaged 3,160 t in 1968- 72. In 1965-76, Ghana's annual catch of Auxis varied from 1,600 to 6,300 t and averaged 3,400 t (Table 26). Japan's resurgence as a fishing power after World War II was not restricted to the Pacific and Indian Oceans. Japanese tuna longliners first made their appearance in the Atlantic in 1957 and by 1961 at least 60 vessels were fishing there for yellowfin tuna and albacore (Borgstrom 1964). In 1962, the Japanese placed pole-and-line fishing boats of 239 gross tons each in Ghana to fish for surface schools of tuna. Yellowfin and skipjack tunas are the main species landed by these Japanese bait boats but Auxis and bigeye tuna are also taken in limited quan- tities (Shomura 1966; Hayasi 1973). Three or four Japanese purse seiners and five to seven pole-and-line (14 in 1972) boats operated in the Atlantic Ocean in 1962-63 and in 1967-72. Tema, with excellent shore facilities, including cold storage, serves as a base for Japanese purse seiners, pole-and-line boats, and longliners (Di Palma 1968). Auxis as well as yellowfin, skipjack, and bigeye tunas are caught along the coast of Africa and in coastal waters of the offshore islands in the eastern Atlantic. In 1967-73, Japanese seiners landed from 177 to 1,256 t of Auxis and averaged 670 t annually (Hayasi 1973, 1974). The proportion of the purse seine tuna catch con- sisting of Auxis, however, was small varying from 3% to 15