Reproduction of the Bluefin Tuna in Captivity - feasibility study for the domestication of Thunnus thynnus

Transcription

1 FINAL REPORT FP 5: Quality of Life and Management of Living Resources Shared-cost RTD Project Project full title: Reproduction of the Bluefin Tuna in Captivity - feasibility study for the domestication of Thunnus thynnus Project Acronym: REPRO-DOTT Contract number: Q5RS Key action 5: Sustainable Agriculture, Fisheries and Forestry, and Integrated Development of Rural Areas including Mountain Areas Sustainable Fisheries and Aquaculture

2 TABLE OF CONTENTS GENERAL INTRODUCTION 8 REPRODOTT PROJECT BACKGROUND 8 Characteristics of the Atlantic Bluefin Tuna (BFT), Thunnus thynnus BFT Stock status BFT farming/fattening current state BFT research status REPRODOTT PROJECT APPROACH 9 Objectives Project Work plan Partnership Project management Summary of the project activities CHAPTER 1: HUSBANDRY EXPERIMENTAL FISH ENVIRONMENTAL CONDITIONS Area description Physical Chemical parameters Introduction Methodology Oxygen Turbidity Description of the current patterns Temperature 1.3 CAGE FACILITIES FEEDING AND GROWTH Feeding Fat content Catotenoids supplementation of broodstock diet Introduction Methodology Results 1.5 ENVIRONMENTAL IMPACT Methodology and studied parameters Sediment quality Benthic communities Rocky substrates Bottom of costal detritus Posidonia oceanica meadows General conclusions CHAPTER 2: HANDLING AND STRESS ON CAPTIVE BFT EMPLOYMENT OF NON-INTRUSIVE METHODS FOR THE ASSESMENT OF BFT SEX AND MATURATION STATUS Introduction Blood Sampling Tagging Muscle Biopsy and Sampling Ultrasound for sex/maturity determination Sex steroids Vitellogenin and Zona Radiata concentrations Dot-Blot (Hot-Blot) Technology 2

3 Test on Southern Bluefin Tuna (SBT) Conclusions Recent further developments for BFT Hot Blots Conclusions and recommendations 2.2 BEHAVIOURAL STUDIES ON BFT SPAWNING Behavioural studies on BFT spawning using Data-Loggers Introduction Methodology Internal Application Data Logger 23 Data Logger 24 External Application Data Logger 24 Data Logger Conclusions Studies on BFT swimming behaviour by Echosounder System Methodology and study material Results and Conclusions Turbidity Bad sea conditions Influence of BFT feeding activity Interferences from Navy ships 2.3 SET UP SUITABLE ANAESTHESIA AND SEDATION PROCEDURES FOR BFT Methodology and study material Results Conclusions 2.4 DESIGNING TRANSPORTATION SYSTEMS FOR BFT AT OFFSHORE AND TO IN-LAND FACILITIES Catching methods Methodology and study material Results and Conclusions Transportation methodologies 2.5 DEVELOPMENT OF A METHOD FOR COLLECTION OF EGGS RELEASED NATURALLY IN CAGES Results in Results in Results in STRESS LEVELS OF CORTISOL, CATHECOLAMINES AND LACTATE IN THE PLASMA OF CAPTIVE FISH Introduction Material and Methods Results Mazarron 23 Lactate and Cortisol Cathecolamines Mazarron Mazarron Conclusions CHAPTER 3: STUDIES ON WILD BFT POPULATIONS SPATIO TEMPORAL PATTERN OF BLUEFIN TUNA SPAWNING IN CENTRAL AND EASTERN MEDITERRANEAN BASED ON GONAD HISTOLOGICAL ANALYSIS Introduction Material and Methods Results Gonad structure and development Ovary 3

4 Testis Spatio temporal pattern of gonad maturation Eastern Mediterranean Central Mediterranean Discussion 3.2. AGE AND GROWTH OF BFT FROM THE MEDITERRANEAN SEA Introduction Materials and Methods Sampling Age and growth estimation Results Discussion 3.3 SIZE AND AGE AT SEXUAL MATURITY OF FEMALE BFT FROM THE MEDITERRANEAN SEA Introduction Material and Methods Sampling Ovary histology and reproductive status Size at first sexual maturity Age determination Results Ovary histology and reproductive status Body weight at sexual maturity Age at sexual maturity Discussion 3.4 REPRODUCTIVE PERFORMANCE OF BFT IN THE EASTERN ATLANTIC AND WESTERN MEDITERRANEAN: INFLUENCE OF SAMPLING GEARS IN THE ASSEMENT OF REPRODUCTIVE PARAMETERS Introduction Material and Methods Sample collection Histology Male histological classification Oocyte and female classification Image analysis and stereology Statistical analysis Results Gonadosomatic indices Histology Stereology Discussion 3.5 DETERMINATION OF STRESS IN WILD BFT POPULATIONS Introduction Methodology and study material Results Season 23 Cortisol measurements Catecholamine values Cortisol and Lactate comparison Season Discussion-Conclusions Wild fish Barbate 24 and Balearic Island Central Mediterranean Baleraic Island DETERMINATIONS OF PLASMA STEROIDS, VTG AND ZRP IN WILD BFT POPULATIONS IN COMPARISON WITH CAGED POPULATIONS Introduccion Materials and Methods 4

5 Steroids Vtg and ZRP Vitellogenin determination in plasma ZRP determination in plasma Results Steroids Wild Fish and Cage Fish 23 Wild Fish and Cage Fish 24 Wild Fish and Cage Fish Vitellogenin and ZRP Levels in Plasma Correlation between Plasma Vtg and Fork Length Correlation between Plasma Vtg and Season Correlation between Plasma Vtg and GSI Comparison of Vtg and ZRP in Wild Fish Mazarron Cages Conclusions CHAPTER 4: SPAWNING ACTIVITY OF BFT IN THE MEDITERRANEAN SEA: WILD VS. CAPTIVE BROODSTOCK ABSTRACT DEVELOPMENT OF TOOLS FOR VERIFYING THE REPRODUCTIVE STATUS OF CAPTIVE BFT BREEDERS Introduction Material and Methods RNA isolation, reverse transcription and amplification of sequences encoding for BFT native GnRH forms and GTHβ-subunits Construction of Pichia pastoris expression vectors Transformation and selection of BFT LHβ and FSHβ recombinant clones Production and purification of recombinant BFT LHβ and FSHβ Antibody production SDS-PAGE and Western Blot analyses Statistical analysis Results Cloning of the cdna encoding for BFT GTH β-subunits Cloning of the cdna encoding for BFT GnRHs Optimization and validation of ELISA to measure BFT LH Development of ELISA to measure BFT FSH Discussion 4.3 STUDYING THE REPRODUCTIVE BEHAVIOUR OF WILD VS. CAPTIVE MEDITERRANEAN BFT Introduction Material and Methods Experimental animals and husbandry Hormonal and histological analyses Results Seasonal profiles of GnRHs and GTHs in Wild and Captive BFT Temporal pattern of gonad maturation Discussion CHAPTER 5: MATURATION IN CAPTIVITY AND SPAWNING INDUCTION USING HORMONAL THERAPIES INTRODUCTION PREPARATION AND ADMINISTRATION OF GONADOTROPIN-RELEASING HORMONE AGONIST (GNRHA)-LOADED IMPLANTS AND INDUCTION OF OVULATION Introduction Material and Methods Experimental animals and husbandry Development of a GnRH Enzyme-linked Immuno-sorbent Assay (ELISA) 5

6 Preparation and validation of GnRHa-delivery systems GnRHa implantation GnRHa-implantation evaluation and gonad sampling Egg collection from cages Statistical analyses Results Development of the GnRHa ELISA In vitro and in vivo release of GnRHa for the developed implants GnRHa implantation Effect of GnRHa implant on plasma GnRHa levels and ovulation/spawning Discussion 5.3 EFFECT OF GnRH TREATMENT ON THE REPRODUCTIVE ENDOCRINE AXIS (BRAIN-PITUITARY- GONADS) Introduction Material and Methods Experimental animals and husbandry Sampling procedures Hormone analysis Statistical analysis Results The effects of GnRHa-implantation on pituitary LH release Correlation between LH levels and successful GnRHa-implantation The effects of GnRHa-implantation on pituitary content of native GnRH forms The effects of GnRHa-implantation on plasma sex steroids profiles Ovarian and testicular development Discussion 5.4 EFFECT OF GnRHa TREATMENT ON GONADAL MATURATION BASED ON HISTOLOGICAL ANALYSIS Introduction Material and Methods Results Histological evaluation of stage of maturation in females Histological evaluationof stage of maturation in males Yolk quantification and stereological analyses Discussion 5.5 EFFECT OF GnRHa TREATMENT ON GAMETE CHARACTERISTICS Introduction Materials and Methods Results Discussion SYNERGY OF THE CHAPTERS AND CONCLUSIONS 194 IMPROVED THE KNOWLEDGE OF THE BFT REPRODUCTIVE BIOLOGY ASSESSED THE CAPABILITY OF BFT TO MATURE AND SPAWN IN CAPTIVITY ASSESSED THE FEASIBILITY TO OBTAIN VIABLE EGGS FROM BFT BREEDERS AND BRING THEM TO HATCHING DEVELOPED HANDLING TECHNIQUES FOR ROUTINE OPERATIONS IN BFT AQUACULTURE GENERAL CONCLUSIONS EXPLOITATION AND DISSEMINATION OF RESULTS 197 DISSEMINATION OF RESULTS Dissemination towards the scientif community Scientific publications 6

7 Contributions to Scientific Conferences and Symposiums International Conferences National Conferences REPRODOTT WWW home page Invitations to annual project meetings Dissemination towards European Commission Dissemination towards ICCAT Dissemination towards fishermen, NGOS and civil society DVD documentary Popular articles and press, radio and television interviews EXPLOITATION OF RESULTS POLICY RELATED BENEFITS 23 LITERATURE CITED 25 References Introduction References Chapter 2 References Chapter 3.1 References Chapter 3.2 References Chapter 3.3 References Chapters 3.4, 3.5 and 3.6 References Chapter 4 References Chapter 5 ANNEX: SCIENTIFIC PUBLICATIONS UNDER REPRODOTT PROJECT 22 7

8 GENERAL INTRODUCTION REPRODOTT PROJECT BACKGROUND Characteristics of the Atlantic Bluefin Tuna (BFT), Thunnus thynnus BFT belongs to the Actinopterygii (ray-finned fishes), order Perciformes, family Scombridae which contains about 32 species and subspecies. There are two species of special commercial interest to fisheries and capture based aquaculture: the Northern Bluefin Tuna with two subspecies: the Atlantic Bluefin Tuna (BFT), Thunnus thynnus thynnus (Linnaeus, 1758), inhabiting the North Atlantic; and the Pacific Bluefin Tuna, Thunnus thynnus orientalis (Temmink and Schlegel, 1844), inhabiting the North Pacific (PBT), and the Southern Bluefin Tuna (SBT), Thunnus maccoyii (Castelnau, 1872) (Ottolenghi et al., 24). The Atlantic subspecie BFT, which is the target specie of the project, is found from Labrador and Newfoundland south into the Gulf of Mexico and the Caribbean Sea and it is also known off Venezuela and Brazil in the Western Atlantic; in the Eastern Atlantic it occurs from the Lofoten Islands off Norway south to the Canary Islands and the Mediterranean Sea. There is also a population off South Africa (FishBase, 22). BFT are considered highly migratory (FAO, 1994) and pelagic species that can be found seasonally coming close to the shore and can tolerate a wide range of temperatures. The fish shoal by size, sometimes together with other tuna or tuna-like species such as albacore, yellowfin, bigeye, skipjack, etc. According with FishBase (22), BFT reach a maximum size of 458 cm Total length and maximum weight 684 kg. Life span is approximately 15 years. BFT live in subtropical waters and it feeds on small schooling fishes (anchovies, sauries, hakes), squids and red crabs. BFT is a major top predator of the pelagic ecosystem of the North Atlantic and Mediterranean Sea that may play key roles in determining food web structure and ecosystem dynamics and whose extinction might induce drastic changes in the functioning and structure of the Atlantic ecosystem, such as cascading effects on the pelagic food web (Essinggton et al. 22; Scheffer et al. 25). BFT Stock status BFT have been captured for centuries in the East Atlantic and Mediterranean waters (Mather et al., 1995; Doumenge, 1998). BFT fishing is mostly carried out using purse seines, long-lines, traps, hand lines and harpoons, drifnets, etc. In 2, The Mediterranean and the Black Sea areas accounted 48% of the global northern bluefin tuna catch (FAO, 22). Because of their flesh quality, BFT are among the most desired and expensive species. The Japanese market ( sushi and sashimi tradition) is the main driving force for the fishery. Currently, BFT has become rare because of massive overfishing (Muus and Nielsen, 1999; Fromentin and Powers, 25). The northern Bluefin tuna fisheries are regulated by the International Convention for the Conservation of Atlantic tunas (ICCAT). In 1996, ICCAT classified the East Atlantic and Mediterranean BFT as being overfished and overexploited. As a consequence, since 1998 regulations and quotas are annually established and revised by this regional management body. The pressure on BFT fisheries in the Mediterranean area has increased considerably over recent years. International organizations such as the Conference on the International Trade of Endangered Species (CITES) have warned about a decline in the abundance of resources and the risk of their exhaustion. As well as NGO and the public have expressed concerns over the status of wild BFT stocks and the threat of extinction of the species due to overfishing (Bregazzi, 26, Atrt, 26) BFT farming/fattening current state BFT has been considered as a good candidate species for marine aquaculture not only because its well known and appreciated by the consumers, and has got high market price and increasing demand; but also due to its very fast growth rate (about 1 times higher than other cultured fish), good ratio of edible meat to body weight (7%) and tolerance to a wide temperature range. Nowadays, BFT are successfully cultured worldwide due to their rapid growth rate, high price and market demand, particularly in Japan. The technology for tuna cage culture has been well developed and industrialized for the Atlantic Bluefin Tuna (BFT), Thunnus thynnus in the Mediterranean countries, for the 8

9 Southern Bluefin Tuna (SBT), T. maccoyii in Australia and for the Pacific Bluefin Tuna (PBT), T. orientalis, in Japan and Mexico. During the last decade, there has been rapid increase in the practice of capture-based tuna farming, particularly in the Mediterranean area. Actually, the farming/fattening of BFT occurs in nine countries: Croatia, Cyprus, Greece, Italy, Libya, Malta, Spain, Turkey and Tunisia. Over 3 registered companies are engaged in this industry, Spain being the leader and followed by Malta, Turkey, Tunisia and Croatia. Mediterranean tuna farming is based on the fattening of wild BFT of different sizes, caught by purse-seine, which are then moved alive to floating cages offshore areas. The fish are then kept in large cages for variable periods, ranging for a few months to years, depending on the farming location and fish size. The actual tuna farming practices is not sustainable neither environmentally nor economical at medium and long term. The first step on the road of a sustainable BFT production is a controlled reproduction. Since larval production technology is underdeveloped, tuna aquaculture is currently dependent on the wild capture of juveniles for production. However, the supply of wild fish is fluctuating in quantity, so establishing hatchery production technology for bluefin and other tunas would be a major step forward in improving sustainability of their aquaculture. In Japan, first natural successful spawning of capture PBT occurred in 1971 at Kinki University and from 1997 PBT eggs were also collected at the station of Fisheries Research Agency in Amami. After more than 3 years of research, Japanese scientists have been able to complete the PBT life cycle, which means that artificially hatched fish spawn and healthy larvae were obtained (Sawada et al., 25). Although, there still exists several difficulties in tuna larval production which include the high mortality during the first 1 days post hatching, their highly piscivorous behavior leading to cannibalism and problems associated with their behavior that it is adapted to the pelagic environment and leads to high mortality by collision with the tanks or net walls (Sawada, pers. Comm..). In addition, the critical factors associated with a low degree of success in the maturation and spawning of captive broodstock remain to be determined. Therefore, the control over the reproduction of captive BFT is not fully achieved. BFT research status For centuries, BFT has been of interest of scientist and even philosophers from the ancient Greek and Latin times. Although, it was since the end of 18 th century, when many European scientists focused on such fascinating species. During the nineties most of the studies were addressed to understand the ABFT biology and the population structure (migration patterns, stock delimitation), thereby providing scientific advice for the stock assessment and fisheries management. In the recent years, apart from the new approaches used to study wild populations (genetics, biochemistry, archival tagging and ecosystem-based ones), an increasing number of researches have been initiated for the domestication of ABFT, coinciding with the quick development of its aquaculture industry in the Mediterranean. Deployment of an increased amount of research efforts has been encouraged not only by the main European countries involved in the BFT fisheries, but also by international management bodies such as the European Commission and ICCAT. In spite of this, there is little scientific information on BFT in comparison to other important fish stocks for UE, such as anchovy, herring, cod, sole or salmon. REPRODOTT PROJECT APPROACH In 23 a 3-year-long research project supported by the European Commission, REPRODOTT (Q5RS ), was initiated aimed at studying the feasibility to achieve reproduction of BFT in captivity conditions. The project draws on the exceptional opportunity to combine forces between an international group of scientists and commercial 'growing out' facilities for BFT to provide an in-depth study of the reproduction of BFT in captivity, thereby providing a strategic feasibility study for the domestication of this species. The controlled reproduction is a crucial step on the road to domesticating a wild species. The project will study the hormonal regulation on gonadal development, maturation of gametes, spawning and fecundity, and 9

10 finally develop suitable methods to control the reproduction of the BFT. The availability of captive broodstock will allow an increase in knowledge from the input of reproductive cues to recruitment determinism in the wild. This will allow better management of the resource in a global sense. Objectives The overall objective of this project is to improve our understanding of the reproductive physiology of BFT as the basis to develop a suitable methodology for the control of its reproduction in captivity, in order to establish a sustainable aquaculture. Four main specific objectives of the project are: 1) To improve the knowledge of the reproductive biology of the species in captivity and compare this with wild populations, in order to develop an aquaculture farming technology for the BFT 2) To assess the capability of BFT broodstock to mature and spawn in captivity 3) To obtain viable eggs from BFT breeders and bring them to hatching must be assessed and found to be feasible 4) To develop handling techniques for routine operations in BFT aquaculture, including setting up safe and effective anaesthesia procedures, designing adequate transportation systems for live fish, developing egg collection devices for cage-reared BFT, and employing non-invasive methods for the assessment of BFT sex and maturation status. Project Work plan The project is structured into nine work packages as follows: WP1. Co-ordination and reporting. WP2. Handling procedures for captive BFT WP3 Maintenance of tuna broodstock in cages WP4 Collection of samples WP5 Histology and gonad morphometry WP6 Sex and maturity determination (brain, pituitary and sex steroids hormones) WP7 Determination of stress in captivity & wild tuna WP8 Hormonal induction of spawning WP9 Gamete characterization and management This following diagram shows the inter-relationship among the different work packages and the objectives targeted by each one of it. 1

11 WP1 Co-ordination WP4 Sample collection Wild and captive WP3 Broodstock maintenance WP2 Handling WP5 Histology WP6 Endocrine WP7 Stress WP8 Stimulation WP9 Gamete quality O1 Scientific Basic Knowledge Technical O4 O2 Spontaneous O3 Stimulated Feasability Seed Production 11

12 Each work package includes one or several deliverables (42 in total) and milestones (29 in total) which are summarised as follows: WP1 Coordination and reporting 3 annual co-ordination meetings 3 annual interim reports and a final report. WP2: Handling procedures for captive BFT Anaesthesia procedures for BFT handling. Suitable transportation systems for BFT. Egg collection methods for caged BFT Non-invasive methods adapted to BFT for sex and maturity assessment. Behavioural studies on BFT spawning. WP3: BFT broodstock management in cages Successful confinement of the broodstock. Reproduction and flesh quality trials. WP4: Collection of BFT Samples Delivery of biological samples from three annual sampling campaigns in both wild and captive fish WP5: Histology and gonad Morphometry Deliver histological and stereological data of wild BFT from three annual sampling campaigns Deiver histological and stereological data of the cage-reared BFT. WP6: Sex and Maturity Determination: WP6.1: Sex Steroids and Vitellogenin Ascertain plasma and tissue sampling procedure and obtain Vtg-rich plasma. Sex and maturity determinations for broodstock and wild fish on three annual seasons. Assays on any induction trials carried out in the last two seasons. WP 6.2: Brain and Pituitary Hormones cdna sequences of BFT GnRHs and GTH β-subunits Quantitative RT-PCR assays for BFT GnRHs and GTH β-subunits gene expression. Optimized and validated ELISA assays for BFT native GnRH forms. Specific ELISA assays for BFT GTHs Seasonal profile of GnRHs and GTHs in wild BFT; gene expression and peptide levels. Correlation between GTH levels and the incidence of successful spawning in treated fish. Plasma levels of conjugated maturation inducing steroids after hormonal induction of ovulation. WP7: Determination of Stress in Caged and Wild Fish Populations Validation of methodology for cortisol and catecholamines as HSP was not used Concentration measurements of stress indicators for caged and wild fish for three experimental seasons. WP8: Hormonal Induction of Spawning Development of specific ELISA for GnRHa. Production and characterization of the GnRHa-implant for tuna. Induction of spawning of BFT in cages (two seasons) WP9: Gamete Characterization and Management Modification of BFT sperm characteristics. Preliminary pilot study report: in vitro induction of maturation and ovulation by steroids and heterologous GTH in seabass and/other species. Description of the main oocyte and egg features in periovulatory period. In vitro induction of BFT maturation and ovulation. Correlations between gamete descriptive characteristics and fertility. 12

13 Partnership The REPRODOTT is a complex project that requires the interaction of BFT farmers and scientists with sound experience in finfish aquaculture, fisheries biology, reproduction and endocrinology. It involves intense field and laboratory work. Nine partners from seven Mediterranean countries were participating in the project: - Spanish Institute of Oceanography (IEO) from Spain was in charge of the project Coordination. Two IEO laboratories were involved. The one located in Murcia was active participant in all workpackages. The other in Malaga was responsible of wild BFT sampling in the Southeast Atlantic and Western Mediterranean. - Tuna Graso, a Spanish private tuna farming company was mainly devoted to the experimental broodstock management. - National Centre of Mariculture from the Israeli Oceanographic and Limnological Institute (NCM- IOLR), was responsible of the Sampling coordination and carry out endocrinological studies on the brain and pituitary. - French Institute for the Exploration of the Sea (IFREMER) was the leader of gametes management studies. - University of Cadiz (UCA) from Spain was conducting the sampling and histological studies of gonad collected in the Western part of the Mediterranean Area. - University of Dusseldorf (UNIDUS) in Germany was the partner responsible for gonadal hormones and stress studies, together with handling activities. - Malta Center for fisheries Science (MCFS) was mainly in charge of wild BFT sampling in their local sea area. - Hellenic Center of Marine Research (HCMR) in Greece was the task leader on hormonal induction of BFT spawning. - Last but not least, University of Bari (UNIBA) from Italy was doing the sampling and histological studies of gonads collected from Malta and Eastern Mediterranean. Project management Taking into account the number and origin of the organisations involved (9 institutions from 7 different countries), their varied expertise in different basic and applied research fields (aquaculture, fisheries biology, reproductive physiology), and the number of work packages foreseen, a good co-ordination plan for the project was considered essential for achieving its objectives. IEO play the role of general co-ordinator of the project with assistance of UNIDUS in performing this task. Each of the other tasks will have a task leader or convenor, who was in charge of co-ordination of the activities carried out in their responsible work package and report its progress to the co-ordinator timely and during the general co-ordination meetings to be held along the project duration. All scientific, financial and administrative aspects of REPRO.DOTT was co-ordinated by Antonio García (IEO), who represent the consortium to the EU. The co-ordinator was responsible for monitoring the expenses and allocation of the budget, organization of meetings, adjustments on the implementation of the project and preparation of annual progress reports. A senior scientist or administrator from each participant was responsible for coordinating laboratory or fieldwork within their group, and for liaising with the other participants. The co-ordinator collected and relayed all relevant information to and from each partner via the responsible scientists. Almost all partners are in charge of one work package, based on their expertise. The activity of the industrial participants was coordinated via the project co-ordinator, trough periodical bi-lateral meetings. The established management committee of REPRO.DOTT is comprised by: 1. Dr. Antonio García (IEO Mazarrón, E) 2. Mr. Antonio Belmonte (TUNAGRASO, E) 3. Prof. Hillel Gordin (NCM-IOLR, IL) 4. Dr. Christian Fauvel (IFREMER, F) 5. Dr. Antonio Medina (UCA, E) 6. Prof. Dr. Cristopher Bridges (UNIDUS, D) 7. Dr. Robert Vasallo-Agius (MCFS, MT) 8. Dr. Constantinos Mylonas (IMBC, EL) 9. Prof. Gregorio De Metrio (UNIBA, I) 13

14 During the second month of the project, a meeting of the participating staff of all teams (including personnel temporarily employed for the project) was held in the laboratories of IEO, with the objective to discuss the organisation of the work and set up standardised methodology, sampling strategies, and distribution of samples. At about 12 monthly intervals, further meetings were held at different locations (France, Greece and Malta) to present and discuss results from the previous campaign and to prepare ongoing field and laboratory work, and drafting Interim and final reports. Apart from that, several informal co-ordination meetings took place coinciding with the summer sampling/induction trials on captive BFT, where almost all project partners were participating. Summary of the project activities 1. Activities addressed to improve the knowledge of the BFT reproductive biology As indicated above, the first objective of the project REPRODOTT consisted in carrying out a detailed study of the reproductive physiology of the BFT, both in the wild and in captivity. Five of the partner institutions of the project, co-ordinated by Professor Hillel Gordin of NCM-IOLR of Israel, carried out analysis on a sample of 6 wild BFT over a period of 3 years, caught by using various methods of capture (purse seine, long-line and traps). In the region of the Spanish South Atlantic and the Western Mediterranean, teams of the IEO and the University of Cadiz were working. In the waters near Malta and the Eastern Mediterranean were teams of the MCFS; while in the Adriatic, Ionian sea, and the waters to the west of Italy was the University of Bari. The last one mentioned also counts on the invaluable collaboration of the team from the University of Istanbul headed by Prof. Isik Oray. With regard to tuna in captivity, all the institutions participated in a sample of more than a hundred fish, carried out in experimental cages in Mazarrón, in the south-eastern part of Spain. All the teams used the same methodology with the sample, so that the results would be comparable. With data concerning the size and weight of the specimens, a morpho-metrical study was carried out, with samples taken from the blood, 14

15 brain, pituitary, gonads, and sexual products; a detailed analysis was carried out by the different partners responsible for the project. Specific methods of analysis of the hormones related with reproduction were developed, as well as a definition of a complete profile of the maturation process of the BFT in the Mediterranean, data that was inexistence up to now. 2. Activities addressed to ascertain if captive BFT were able to mature and spawn naturally Finding out if the tuna held in captivity for various years could mature and spawn eggs in a natural way, was another fundamental objective of the project, because if this was the case it would open the doors to its future rearing in captivity. Thus, at the beginning of the project, a stock of some wild breeders was established. The company, Tuna Graso, a partner in the project, was given the task of the fish maintenance, supplying them with the best possible food and carrying out a control of the different physical-chemical parameters (temperature, salinity, dissolved oxygen currents regime, etc). At the same time, an annual study of the environmental impact of the cages was carried out. One of the ways to test if the tuna were spawning eggs consisted of observing their behaviour during the months of June and July by the presence of observers watching closed to the cages at sunset to see if the tuna were coming to the surface or to find out any unusual movements, in such a way that would indicate a courtship phenomenon or the emission of eggs and their fertilization. Another system for observing their movements, pioneer for this purpose, was the positioning of echo-sounding equipment. Moreover, a number of fish had electronic recording tags implanted (data-loggers), which stored data about their position and temperature every certain period of time. To find out if the fish had emitted eggs, various types of collecting nets were designed and tested in the fish cages. In addition, nets and small collectors were passed daily over the surface and around the cages so as to collect possible eggs. Finally, a number of fish were sacrificed to carry out histological and hormonal analysis to check if they had reached the necessary maturity for reproduction. 3. Activities addressed to check the feasibility to obtain of BFT viable eggs and larvae In order to induce and promote the BFT maturation and spawning within the cages, Dr. Constantinos Mylonas from HCMR, Greece with the help of other scientists, made a suitable hormonal implant for such a large fish species. Thereafter, during the summers of 24 and 25 several trials were conducted on the use of specific designed darts containing the hormone. Those darts were propelled through a submersible spear gun by professional divers from Tuna Graso. Few days later, fish were killed to assess the effect of the treatment by the different project participants teams 4. Activities addressed to develop handling techniques At the same time as all these previous activities were being carried out, the IEO team and UNIDUS, with the participation of the rest of the teams, developed techniques for handling this species. This was necessary not only because of their large size, but also because of their swimming speed and their special sensitivity to being handled, which can result in causing them grave injuries and even death due to the stress they suffer. The first step was to develop techniques for anaesthetizing them. Another problem was how to capture the individual tuna to be sampled and /or to be treated hormonally, without disturbing the rest of the fish. Other approach was to develop ways of assessing the sex and maturity stage of the tuna without causing them any damage, such as echographies and muscular biopsies. A more detailed explanation about these activities: material and methods employed, results obtained and its discussion, as well as the main conclusion achieved can be found in the following 5 chapters. In order to be more clear and comprehensive, these have been split into a consolidated way having the main titles as follows: Chapter 1: Husbandry Chapter 2: Handling and Stress on Captive BFT Chapter 3: Studies on Wild BFT Populations Chapter 4: Spawning Activity of BFT in the Mediterranean Sea: Wild vs. Captive Broodstock Chapter 5: Maturation in Captivity and Spawning Induction using Hormonal Therapies 15

16 CHAPTER 1: HUSBANDRY 1. 1 EXPERIMENTAL FISH Adult BFT (over 12 individuals averaging 4-6 kg mean body weight) were obtained by TUNA GRASO in (Year 1), captured by the purse seine fleet operating in the Mediterranean, then transported to Spain in floating cages at a very low speed, and transferred to holding cages. Half of the specimens (64 BFT), caught in the wild in summer 22 and reared on cages since then, it were stocked on a floating cage as early as February 23. The other half (another 64 BFT) were caught in early summer 23, stocked in a big cage (5 in diameter) and transferred to two 25 m cages in October 23. At the moment of the transfer to the cage some BFT were tagged for an easy visual identification, and simultaneously a muscle biopsy was obtained from each fish. IEO and Tuna Graso conducted these operations with the assistance of UNIDUS and UCA for tagging and muscle biopsies collection. UNIDUS sexed the tuna from steroid hormone determination on muscle samples. Part of these tuna was maintained for three years. A thorough survey was established to detect presumable reproductive behaviour and/or spawning in the cage. Environmental water parameters in the cages were measured throughout the project duration. Feeding behaviour and occurrence of diseases were also noted from stocked BFT. An environmental impact assessment and yearly survey on the sea bottom area under the experimental cages accompanied the project. 1.2 ENVIRONMENTAL CONDITIONS Area description The coastal bay where the experimental cage facilities are, is situated at the occidental extreme of Cape Tiñoso, (Mazarrón Bay, Murcia, South East Spaint). It is a coast facing south, so that the wind and waves exposures are predominant in the 2nd and 3th quadrant. It is characterized by a low coast, with some middle-low cliffs and a shallow bathymetric profile, which give very good conditions for the growth of the seagrass Posidonia oceanica and also has soft bottoms (sand and pebbles), with prominent beaches. Many coastal watercourses end in Mazarrón bay or inlet Physical-chemical parameters Introduction The general pattern of the water masses on the coast of Murcia (Coastal mass= Mediterranean, warmer and salty and the Atlantic mass further away from the coast) is broken near Tiñoso cape, since right here upwelling it has been detected as deep waters, forced by the presence of submarine valleys that cross the continental slope rises to the surface. This presence of colder waters near the coast induces a higher content than the general oligothrophic nature of the area Methodology A current-meter RCM9, which was anchored at 12 meters depth near the experimental cages, was recording hourly water parameters such as direction and velocity (intensity) of water currents, salinity (conductivity), turbidity, oxygen concentration, temperature and pressure in the area where the experimental cages were located. These data were downloaded every two months until August 24, when the equipment was probably stolen. Besides, two max-min thermometers were placed at two different water depth (6 and 12 meters) Oxygen 16

17 marchapril may-june july-sept sept-nov 23 nov-dec 23 janmarch 24 marchmay may-aug Maximum concentration (mgrs.) Minimum concentration (mgrs) Mean concentration (mgrs.) Fig. 1.1 Oxyen concentrations recorded at the BFT cage facilities during the experimental period Figure 1.1 shows the maximum/minimum oxygen concentration (in milligrams) during the period As it can be seen, the waters where the study was carried were very well oxygenated. After August 24 no data were recorded since the current-meter disappeared Turbidity marchapril 23 mayjune 23 july-sept 23 sept-nov nov-dec janmarch 24 marchmay 24 may-aug 24 Maximum concentration (NTU) Minimum concentration (NTU) Mean concentration (NTU) Fig. 1.2 Turbidity recorded at the BFT cage facilities at the experimental period. Turbidity values were also at low levels during the experiments. Again no data were recorded after August 24 because of the disappearance of the current-meter. Waters that surround the area where the facilities have been located are characterized by being oligotrophic, without many nutrients, transparent and for not having any contribution from city or industrial centres. The only changes regarding nutrients or turbidity are induced by natural causes, the upwelling of deep waters, vertical exchanges and periodic inputs of land material Description of the current pattern In the West of Tiñoso Cape (Azohía), the frequent current pattern direction is NW-SE (parallel to coast), where the predominant current is to the SE, with high speeds and a duration of approximately 14,7 days (lunar cycle), which is interrupted by a change in the current pattern to NW, with less intensity and at the same time less variability. The following diagrams show the speed of sea current and the rose of directions (Figs. 1.3 and 1.4) 17

18 Fig. 1.3 Speed of the sea current (cm/sec) versus the two main geographical coordinates axes (North-South and East-West) Fig. 1.4 Rose of sea current directions indicating the main orientations by the length of the arrows Temperatures. Fig. 1.5 shows the temperatures recorded during the experimental period (23-25) at 6 and 2 m depth.the temperatures ranged from 13-15ºC in winter times to 25-27ºC in summer. Minimum temperature recorded was 12,47ºC (March 25) and maximum 27,7ºC (August 23). No significant differences were observed between the temperatures at two depth from October to April, but this month the appearance of a thermocline denoted marked differences, particularly in August where a difference of up to 5 degrees between the two depths was observed. 18

19 TEMPERATURE CHANGES (ºC) TEMPERATURE mar-3 jun-3 oct-3 ene-4 abr-4 ago-4 nov-4 feb-5 may-5 sep-5 MONTH MAXIMUM -6 meters MAXIMUM -2 meters MINIMUM -6 meters MINIMUM -2 meters Fig. 1.5 Variation of seawater temperature at the BFT cage facilities during the experimental period, recorded at 6 and 2 meters depth. 1.3 CAGE FACILITIES To carry out the research activities on captive BFT foreseen in the project workplan, two 25 m in diameter experimental floating cages of the Polar Circle type, were anchored off Mazarrón (Murcia, Spain), inside the best area available inside the limits of the tuna farming public sea concession avoiding the exposure to swells or storms. Another cage of 5 m in diameter, non-budgeted on the project, was provided for free by TUNA GRASO Company. Fig. 1.6 BFT cage faciltties consisted on two 25 meters in diameter experimental cages (smaller rings) and one 5 m (large ring). 1.4 FEEDING AND GROWTH Feeding From Monday to Saturday a ship with its crew (mariners and skipper) and a diver (Tuna Graso staff) went to the experimental cages to feed the BFT broodstock. Food was administrated till satiation, stopping the food allowance when divers notice that BFT showed no interest for the food and noting this downit. The daily amount and type of raw fish employed for feeding, was noted. Food consisted on Atlantic mackerel (Scomber scombrus), sardine (Sardina pilchardus), round sardinella (Sardinella aurita), herring (Clupea arengus) and squid. Feeding during the project, including quantity and kind of food administered every month is recorded in the figure 1.7 and Table

20 FEEDING DURING PROJECT 7 6 QUANTITY OF FOOD (Kg) feb-3 abr-3 jun-3 ago-3 oct-3 dic-3 feb-4 abr-4 jun-4 ago-4 oct-4 dic-4 feb-5 abr-5 jun-5 MONTH Fig. 1.7 Quantities of raw fish supplied to the BFT experimental fish during the project It must be pointed out that the quantity of food is decreasing from the beginning because of the first trials and casualties. It increases in late September of 23 because it is when the new tunas are caged, decreases again in February-march of 24 due to trials with data loggers and bad weather, after that, the appetite seems to rise and after June of the same year it decreases again due to the induction trials and after that, also because of slaughters of BFT specimens for sampling. Table 1.1 Total amount of different raw fish used as food for the experimental BFT during the experimental period. Total kilograms Scomber japonicus Scomber scombrus Sardine Herring Sardinella aurita Squid Alache Fat contents Variations in fat content might be related to the the broodstock reproductive status or stress that they are exposed to. During 23, mean values for fat content during May sampling, (tunas from cage 1), is 16.33%. During June sampling the mean value decreases as far as 9.15%, being most of the tunas from cage number 2. Samples obtained from tunas from cage number 1 had a higher concentration. Samples from the tuna sacrificed on the 15th of July in cage number 1 show a fat level similar and even higher (17%) than the previous one, while samples from the tunas from cage number 2, sacrificed or dead in July (from 28 to 3) show a decrease as far as 5.31%. Data from 24 and 25 are summarized in table 1.2. Table 1.2. Fat content of the meat of BFT specimens sampled in 24 and 25. Tuna Samples 24 Tuna Sample Fat Content Tuna Sample Fat Content ZZ 13 4 % ZZ % ZZ 14 1,2 % ZZ 114 6,8 % ZZ % ZZ 115 9,6 % ZZ 16 11,2 % ZZ 116 9,2 % ZZ 17 8,4 % ZZ 117 5,2 % ZZ 18 6 % ZZ 12 6,8 % ZZ 19 6,4 % ZZ 121 6,8 % ZZ 11 5,2 % ZZ 122 8,8 % ZZ 111 5,2 % ZZ 123 2,4 % ZZ 112 9,6 % Tuna Samples 25 2

21 Tuna Sample Fat Content Tuna Sample Fat Content ZZ 21 7,6 % ZZ 221 No data ZZ 22 13,2 % ZZ 222 4,8 % ZZ 23 8,8 % ZZ ,2 % ZZ 24 2 % ZZ 224 1,4 % ZZ 25 9,2 % ZZ 225 2,4 % ZZ 26 2 % ZZ 226 2,8 % ZZ 27 8,4 % ZZ 227 4,4 % ZZ 28 4,8 % ZZ 228 7,6 % ZZ 29 ZZ % ZZ 21 7,6 % ZZ % ZZ % ZZ ,4 % ZZ 212 6,4 % ZZ 232 1,4 % ZZ 213 1,4 % ZZ % ZZ 214 7,6 % ZZ 234 6,8 % ZZ ,6 % ZZ ,8 % ZZ % ZZ 236 8,8 % ZZ 217 No data ZZ % ZZ % ZZ % ZZ 219 6,4 % ZZ 239 No data ZZ 22 5,2 % ZZ 24 1 % Carotenoid supplementation of broodstock diet Introduction The attractive and bright pink or reddish colour of flesh from some marine finfish likes the Atlantic salmon (Salmo salar) or trout (Oncorhynchus mykiss) is a natural and characteristic attribute that the consumers associate with quality and trust. This colour is a consequence of the depot of significant amount of carotenoids in muscle tissues (5-1 µg/g; Page y Davies, 26). However, the muscle is not the main depot tissue in fish for carotenoids except for the above mentioned species, but the skin, gonads and sex products (eggs, sperm) and liver. Through supplemented diets with carotenoids is possible to increase notably the amount of pigment. In fact, in the formulation of commercial diets for fish farming usually employ the carotenoids astaxantin and cantaxantin. Besides the colour, carotenoids may produce a number of other bioactive beneficial properties on the fish such as immune-modulators (Amar y col., 24), or the improvement of reproductive performance (Agius y col. 21; 22). These pigments are precursors of vitamin A. Thus, once ingested in the diet they can be transformed through a number of metabolic ways into this essential vitamin related to several physiological activities (Matsuno, 21). This approach might be of interest of tuna farming industry, not just for the reproduction itself but to improve the meat color, and therefore selling price too. Capsorbin and capxanthin, the main carotenoids on paprika, may play a role in the reproductive axis, improving the reproductive performance. This action could be related to the antioxidant or provitamin properties of these carotenoids Methodology From April 25, BFT broodstock from the two experimental cages were fed with bait fish with their stomach filled with a paprika paste. The amount of paprika powder was 2% of daily food allowance, in accordance with MCFS suggestion. BFT liked this food and even increased their appetite and food consumption. Since the work of supplementing the bait was labour intensive and time consuming, the paprika paste was soon after replaced by another presentation of paprika, which was oleo-resins. The latter is a concentrate that could have much more xanthophylls than the powder. Then, the amount to be introduced on the bait fish could be reduced by using the paprika oleoresin as a replacement for the powder. To ascertain the actual concentration of carotenoids in the oleoresin, the collaboration of the Department of Food Biotechnology from the Fat Institute belonged to the Higher Scientific Research Council (CSIC) to 21

22 make the biochemical analyses was sought. This is a Spanish public research centre working in this kind of food ingredients for several years Results of the analyses indicated that paprika oleoresins contained 27 times more capsorbin and 17 times more capxanthin than the powder one (Table 1.3). Therefore, the amount of oleoresin as supplement of the diet was reduced from 2% to just.4%. In practical terms, instead of the daily using of 1 kg of paprika powder into the stomach of about 1 mackerels, it was used 2 gr into 2 bait fish. Table 1.3. Quantitative analyse of the different carotenoids in paprika oleoresins and powder. Paprika Oleoresin Paprika Powder Times more % capsorrubin 5293,15 5,86 capsorrubin 198,89 5, violaxanthin 1283,99 1,42 violaxanthin 153, 3, capxanthin 29729,3 32,89 capxanthin 1766,93 45, Cis-capxanthin 24364,66 26,96 cis-capxantin 546,94 13, Capsolutein 5975,3 6,61 capsolutein 321,39 8, zeaxanthin 4552,96 5,4 zeaxanthin 33,47 8, Cis-zeaxanthin 585,1 6,47 cis-zeaxanthin 84,88 2, beta-criptoxanthin 5533,7 6,12 beta-criptoxanthin 274,6 6, beta-caroten 77,75 7,82 beta-caroten 246,36 6, Cis-betacaroten 19,24 1,21 cis-betacaroten,52, Total 938,47 Total 3923, Red carotenoids 59386,84 Red carotenoids 2512, Yellow carotenoids 3993,63 Yellow carotenoids 141, R/Y 1,92 R/Y 1, Results After four months of food supplementation with paprika, BFT samples of muscle and gonads were obtained in July and delivered to the Fat Institute to obtain their carotenoids profile. Muscle samples of 19 BFT were processed to determine the carotenoids content after the supplemented diets with paprika powder and oleoresins. The samples were homogenised, filtered and the lipid fraction extracted. Thereafter, they were analysed by UV-visible spectrophotometer to determine qualitatively and quantitatively the profile of carotenoids into the sample. Results indicated that there were no carotenoids in the muscle, since at its characteristic absorption range (Fig. 1.8) of 4-5 nm the curve was flat. On the contrary the main absorption area happened in the range between 3 and 4 nm, which means the presence of polyunsaturated lipids (Fig. 1.9). This is consistent with the previous studies indicated above, which seems to indicate that muscle is not the main tissue for the depot of carotenoids. On the contrary, the muscle samples smell typically of paprika. Although this organoleptic flavour could have been transferred to the flesh because paprika contains high values of this aroma. 22

23 Fig. 1.8 Typical UV-visible absorption spectrum of carotenoids 2,6 2,2 1,8 Unidades de Absorbancia 1,4 1,,6,2 -, Longitud de onda (nm) Fig. 1.9 UV-Visible spectrum of the lipid extract of BFT muscle tissue In order to check if the paprika supplemented diet might increase the carotenoids content in other tissues such as gonads, samples from 4 BFT from the same stocks were processed and analysed. This time results indicated the presence of significant amount of carotenoids. The nature and exact quantity of this special kind of carotenoids are being determined by the Fat Institute by HPLC-MS and EI-MS analyse methodologies on the lipid extracts. The fact that the higher maturity degree achieved by the BFT broodstock this year seems to indicate that the paprika supplemented diet may play a positive effect on their reproductive performance. Although, further experiments should be conducted to check this hypothesis. 1.5 ENVIRONMENTAL IMPACT Methodology and studied parameters Sediment quality: The granulometry, organic matter, organic carbon and microbiology (Faecal and total coliforms and faecal streptococcus) were measured Benthic Communities: 23

24 Rocky substrate It was determined by means of scratches of 2 x 2cm, in and 1 metres depth. for all samples. The algal community were studied using these samples Bottoms of coastal detritus By means of box-corer (PVC cylinders 5 cm. in diameter). Sediment samples were taken to study the meiofauna. By means of the collection of squares of 2x2 cm from the first 5 cm of the sediment, polychaeta were studied as the more representative groups. Posidonia oceanica meadows - Sheaf density: It indicates if the seagrass beds are in progression, regression or in an stationary state. - Sheaf characteristics: Reduction of the foliar growth due to the limitation of any resource (Light, nutrients, and so on) or the lost of biomass because of the action of secondary factors (herbivorous), ends with the modification of sheaf characteristics what is named as sheaf size. - Abundance and nutritional composition of epiphytes (algae, hydrozoans and bryozoans): In the presence of additional dissolved nutrients, epiphytes that grow on the phanerogams leaves are able to develop fast biomass increases which stimulate herbivorous activity. Epiphytes are able to undergo a fast nutrients assimilation which will involve changes in the elementary composition of its tissues. Therefore, the epiphytes nutritional composition can be used as indicator of a probable increase of the nutrient availability in the environment General conclusions As regards to organic material it is shown that levels are similar throughoutl the study period and far from those from areas with external inputs caused by aquaculture activities or the proximity of macrophitic dense seagrass beds. The polychaetes keep a faunistic composition which can be considered as normal and are a well structured community but there are differences between the control points under the farm and the others, since under the cages of the farm an effect with time is clearly shown. During past years the detected alteration in diversity and richness indexes in farming periods has been alternating with recuperation processes through the period without farming. Algal community in the rocky infralittoral, keep the tendency observed previous years to lose the wooded stratum in favour of a grass component because of the incorporation of opportunistic filamentous algae. Tendency of regression of Posidonia oceanica near the farming installations is confirmed because of the regression of the lower limit, the density reduction and the coating of the surviving deep seagrasses. Generally, none of the detected alteration can be attributed to the research cages for REPRODOTT project, since they had been previously identified in the Vigilance Environmental Plan for the tuna farming operations and so that related on the whole activities inside the sea concession. Some months after complete cessation of activity in Tuna Graso s concession, some parameters some showed improvements and recovery signs in the communities affected by the aquaculture activity. 24

25 CHAPTER 2: HANDLING AND STRESS ON CAPTIVE BFT 2.1 EMPLOYMENT OF NON-INTRUSIVE METHODS FOR THE ASSESSMENT OF BFT SEX AND MATURATION STATUS Introduction The specific physical characteristics and behaviour of BFT (large body size, fast and continuous swimming, high sensitivity to catch and transport operations which usually cause physical damage such as skin lesions or abrasions of fish and mortality) require the development of new handling methodologies and strategies. The high commercial value of the fish will force the farmer to reduce, as much as possible, the risk of losses during manipulations. These restrictions therefore impose constraints on the study of the reproductive physiology of the fish. One of the crucial problems in modern aquaculture is the determination of the sex of the brood stock and also the sexual maturity, which in the Bluefin Tuna, Thunnus thynnus is reached after 3 years in the Eastern Atlantic stock. This is important in determining the sex ratio of the brood stock for stocking purposes and also for separating individuals apart for hormonal induction experiments. These types of determination are routinely carried out with many of the marine aquaculture species such as sea bass and sea bream by anaesthetizing the fish and carrying out a biopsy for gonadal material or even using no anaesthetics and taking a rapid blood sample or gonadal biopsy. A new non-intrusive method for the assessment of gonadal development consisting of the use of ultra-sonography has proved successful in salmonids, pacific herring, lampreys, groupers, sturgeons and Atlantic halibut (Martin et al., 1983; Reimers, et al., 1987; Bonar et al., 1989; Mattson, 1991; Mazorra de Quero et al., 1999; Martin-Robichaud and Rommens, 21; Moghim et al., 22; Maeva et al., 24; Whiteman et al., 25; Wildhaber et al., 25 ) In order to carry out research in reproductive physiology studies, it is necessary to collect tissue samples from fish without killing them in order to determine the sex and maturity state of the fish. The aim is to get the samples in a safe way, reducing stress and injuries to the fish. Sex and gonad maturity stage in fish studies can be determined by gonad biopsy and this has been the method of choice in most aquaculture studies. In many previous studies plasma samples are being taken, puncturing the branchial arches, the caudal vein or the heart of fish with a syringe is usually employed to collect blood samples for plasma hormonal level quantification. These techniques may be adapted if necessary to the specific physical characteristics of BFT, by using modern techniques maturity and sex can be determined from a muscle sample alone (Coward and Bromage, 2; Susca et al., 21; Bridges et al., 21;Bridges et al., 23). Further studies have also been made one determinations made in mucus taken from the fish. Using a combination of these methods the present study outlines the methodology and adaptations used to obtain samples and the reliability of the use of the biomarkers for sex for example the SSF (sex steroid formula) or the use of the Dot Blot involving antibodies for vitellogenin or Zona Radiata Protein Blood sampling Besides the collection of muscle samples for hormones levels measurements, blood sampling was also considered essential for such aims. An anatomical survey aimed to localise putative sampling points for blood collection was carried out. Although the bulbus arteriosus from the heart was considered as a suitable place for blood sampling, the specifical BFT peripheral blood vessels system, the so-called rete mirabilis, was recognised as an easy and quite safe sampling place. Nevertheless, some concerns about the possibly of collateral damages on important organs, such as lateral line, when using this techniques were pointed out. 25

26 A B C Fig. 2.1 Blood sampling procedures in BFT: A) either from the heart (bulbus arteriosus), or B) below the lateral line. C) Masscollection of blood by cutting the lateral line at the peripheral blood vessels location Tagging For the individual identification of fish colour or number coded tags have evolved and also electronic chips or PIT Tags (Passive Integrated Transponder) through which individual identification tags are coded for electromagnetic reading. In the present study a visual tag was conceived which could be identified both underwater and on the surface. This has been combined with an applicator system, which allows the taking of a muscle sample at the same time as applying the identification tags. In Fig. 2.2 different tag designs are presented, one a non-floating tag and the other a floating tag. These were designed as proto types at the beginning of the study. Later Fig 2.3 different tag designs were tested and also included in the induction tags which were both colour coded and had indicator zones to show depth of penetration. The plastic arrow head anchors were obtained from the Floy Tag company and were modified to hold a PIT tag. Figure 2.4 shows the combination of the arrowhead tag which had been modified to hold a PIT tag and the muscle biopsy needle. Fig.2.2 Two types of tags used for visual identification of fish. A: the first simple prototype and B: the modified floating system. 26

27 Fig. 2.3 Various prototypes of identification tags for use on a hand-held implanter or spear-gun During 23 in the initial stage the tags shown in Fig 2.2 A were used together with the Jab-Stick approach. The tagging results were mixed and although the tags and sampling could be carried out under water successfully the first prototype caused skin abrasion and tag loss due to braiding of the monofilament. The floating tag was more successfully and the position and retention was confirmed by underwater photography. From the biopsy samples a number of parameters were measured and it was found that a combination of steroid analysis together with the presence or absence of significant amounts of vitellogenin were good indicators of sex even in the early part of the reproductive season. A B C D. E C 27

28 Fig. 2.4 Tagging procedure and shortcomings. A) Taggs labelled ready-to-use; B) Taggs fitted to the jabs-stick; C) tagging BFT by scuba divers; D) BFT tagged; E) Skin damages caused by the first tag prototype; F) detail of the skin lesion caused by the frequent scrapping of the tag against fish body Muscle Biopsy and Sampling The muscle biopsy needle was affixed to a jab-stick and used by a diver to implant a visual marker and take a muscle biopsy sample (2mg) at the same time in specimens of Bluefin Tuna held in the facilities of Tuna Grasso in Mazarron, Spain. Within the framework of a previous study it has been possible to develop a methodology for sex and maturity determination from both plasma and muscle biopsy from the BFT Ultrasound for sex/maturity determination Some trials on the use of an ultra-scanning machine (Honda HS 2V) for a non-invasive sex and maturity assessment on BFT were carried out (fig, 2.5). The employment of some multi-frequency probes, both linear and convex, was tested. Several progress on getting skill and experience required to accurately ascertain the sex/maturity of BFT were achieved. The characterization of ultrasonic images of male and female gonads is underway. Although the fish and gonad size (over 5 kg), and the thickness of abdominal muscling caused some difficulties. A B Fig. 2.5 Ultra-scanning machne (A) and ultrasonic images of BFT gonads (B) Sex steroids 11-Ketotestosterone (11-KT), Testosterone (T) and 17ß-Estradiol (E 2 ) were extracted with Dichloromethane and the concentration measured by ELISA as described by Cuisset et al. (1994). In our laboratories we have established ELISA's with a sensivity of 1 pg/ml for 11-KT and T and 3 pg/ml for E 2 in the assay. We converted the different sex steroid concentrations into a mathematical relationship applying the Steroid Sex ratio Formula (SSF): [(E 2 /11-KT)]/[T]. We plotted SSF vs. T concentration on a logarithmic scale from Spanish BFT plasma and muscle samples (Fig, 2.7 and 2.8 respectively). Testosterone representing a marker for sexual maturation for both males and females. Fig. 2.6 The combination adapter for muscle biopsy tissue sampling and electronic and visual tag application. 28

29 1 1 a 1 y = -,66x +,49 r 2 =,2125 log (Plasma SSF) Females Males Regression y = -1,19x - 1,42 r 2 =, log [Plasma T] Fig. 2.7 Results of SSF using plasma samples from Spanish BFT fish and determining all parameters 1 b Females Males Regression log (Tissue SSF) 1 1 y = -,81x +,19 r 2 =,6336 y = -1,7x +,92 r 2 =,6839,1 1 1 log [Tissue T] Fig. 2.8 Results of SSF using muscle samples from the same Spanish fish and determining all parameters. It can be noted that the measurements made with the plasma samples give excellent separation of both males and females in a matter of what the testosterone concentration is. This type of result has been found in previous measurements made on Mediterranean tuna and is backed up by recent measurements on the Mazarron fish. With the tissue sampling there are areas of overlap as can be seen from the figure above. The actual time of sampling for plasma i.e. seasonality does not appear to be important. For the measurements with tissue this may be a problem when the fish are in a non-reproductive state Vitellogenen and Zona Radiata Concentrations The presence of Vitellogenin, a lipoprotein precursor for the yolk protein, or Zona Radiata (ZRP) can also be used to determine sex and also sexual maturity stage. This involves the use of two specific ELISA determinations developed in our laboratory. The concentration thresholds can be set for males and females. In general the concentration of vitellogenin in plasma of males is extremely low at normally less than.5 mg/ml. Using this threshold (fig. 2.9) then approximately 9% of the examined fish could be sexually identified correctly. Approximate to the 8% of the females were classed as males and only 3% of the males classified as females. This can be explained by extremely low vitellogenin values in immature female fish and higher vitellogenin values in some males which may be due to endocrine disruption. 29

30 % Female fish classified as males 8 % False Sex Determination 3 % % Male fish classified as Females Correct Sex Determination 89 % PLASMA Fig.2.9 Percentage correct or false determinations using Plasma Vtg thresholds. A similar analysis using muscle samples (fig. 2.1) together with the ZRP and vitellogenin was not so accurate for male fish with a 28% error but this was using combination results of ZRP and Vtg. If ZRP was used separately the results were better. The problem here is deciding on a female threshold. If the comparative sample are all immature then the muscle levels will be extremely low as shown below and therefore give greater error. If the females are approaching spawning then the threshold can be set higher with a better % of correct determinations. % Male fish classified as Females 3 % 28 % False Sex Determination % Male fish classified as Females Correct Sex Determination 69 % MUSCLE Fig.2.1 Percentage correct or false determinations using muscle Vtg thresholds. In the last part of this study Mucus samples were examined (fig. 2.11) an again gave a 7 % absolute accuracy and roughly 15% either way for males or females. 3

31 % Female fish classified as males 13 % False Sex Determination 17 % % Male fish classified as Females 7 % Correct Sex Determination MUCUS Fig.2.11 Percentage correct or false determinations using mucus Vtg thresholds. All of the studies shown above are our empirical ones where the threshold levels have been taken to be the mean level measured in female fish. It is clear that this threshold changes with seasonality and sexual maturity of the fish. An examination of the raw data on a case-by-case basis gave far better results. However the factor of uncertainty will always be present, although the combination of the SSF together with Vtg and ZRP are a powerful tool Dot-Blot (Hot-Blot) Technology A modified Western-Blot analysis known as Dot- Blot was carried out as described by Susca et al.(21). The primary antibodies for the Dot-Blot analysis were the same rabbit anti-bft vitellogenin and rabbit anti- Bluefin tuna Zrp sera used for the immunohistochemistry. The secondary antibody used was an anti-rabbit- IgG alkaline phosphatase conjugate (Sigma Aldrich, Taufkirchen). As a negative control an assay buffer (.1M PBS, ph 7.4,.5% Tween-2, 27mM KCl) was used. The positive control comprised of Vtg or Zrp standard at a concentration of 1 _gml 1.Each application spot represents 1µl of control, standard plasma or muscle/gonad extract Tests on Southern Bluefin Tuna (SBT) In the Southern Bluefin Tuna, SBT, (Thunnus maccoyii) fisheries around the coasts of Bali Indonesian fish are caught on long-lines and are normally gutted at sea. This therefore gives rise to problems of sex determination. To date manual inspection of the genital tract looking for remnants of Gonads or Testis has been used to sex fish. The present tests were carried out to establish the suitability of a Dot Blot test using antibodies generated from Bluefin tuna, BFT (Thunnus thynnus) for the lipoprotein vitellogenin and also the eggshell protein Zona Radiata. The use of these two biomarkers at the required concentration should enable one to distinguish between males and females using simple techniques, with a little equipment. The present study should indicate the suitability of the Antibodies chosen and their response in Southern Bluefin tuna and also give some idea of the Tests ability to determine between males and females. The tuna caught south of Bali are normally in the spawning areas and therefore concentrations of and vitellogenin and Zona Radiata should be high. Assessments were made on both muscle tissue and also pieces gonad taken on landing. 31

32 Table 2.1 Biometric Data of SBT caught near to Bali, Indonesia Small 1x1 centimetre sample of muscle or gonad were placed in a micro homogeniser (see Figure 2.12 below) and homogenised with a one ml of PBS buffer. Fig Micro homogeniser system made out of glass for muscle or gonad samples. Results of scans of Dot-Blot with Vtg are shown in the below figures: 32

33 Fig Scan of Dot-Blots results made using Anti-ZRP antibody It is clear that antibodies raised for BFT Vtg and ZRP are quite capable of giving an adequate signal for SBT and probably ZRP will be the preferred antibody be to use. On a whole the signal from gonad samples were a magnitude higher than that obtained from the muscle samples for both Vtg and ZRP. It would therefore be possible to assay any remnants of gonads ovaries or testes which were still attached to the reproductive openings in the fish. Again the subject use of these characteristics have been used by the field observers so there would also be enough material to make an homogenisation from and depending upon whether it was a male or female then the signal obtained would be appreciable. The colour discrimination window for males is small using the first trial experiments but even so the muscle samples from males were distinctly lighter than those from the females for a non-trained observer. This was also the case with the ZRP samples. Basically the concentrations of the 2 antibodies used and the dilution of the tissue sample will give us a threshold which will enable us to determine females accurately. This would mean setting the discrimination level of Vtg analysis that it only gives a colour response above 1mg/ml which would be greatly exceeded by all spawning females. All males would fall well below this value and hence give no colour reaction. Similar thresholds can be determined for ZRP and the Hot-Blot assay adjusted accordingly. Again it must be emphasized that this was a pilot study using the single spot Hot Blot Eppendorf system. A more sophisticated set up using full sheets/petri dish size can be used to analyse up to 2-5 spots at a time. This will depend on your sampling protocol and needs etc. There are some problems with the method which involve non-specific background binding of Tuna proteins. This can be accounted for and the effect reduced by using immuno precipitation of the antibodies with control plasma. Again balancing of antibody concentration gives the necessary results and discrimination window. The genes for generating Vtg and ZRP can be switched on in male fish via Endocrine disruption this can lead to incorrect identification of female fish but so far this has not been observed in and Tuna species with any certainty. Most results to date on the Pacific Ocean Tuna have been negative Conclusion The use BFT antibodies are quite adequate for identification of male and female SBT. The use if of the Hot Blot in field conditions should give accurate results in terms of the identification of female fish. The Hot Blot method will be developed further for BFT. Its use for SBT will only require minor modifications and 33

34 adjustments of the antibody concentrations. The method itself will be extensively worked on in Düsseldorf to provide a clear cut all-or-nothing response in BFT. The generation of specific antibodies for SBT can also be achieved in either Dusseldorf, if material is available i.e. plasma from mature females (VTG) or mature gonads (ZRP). The use of the sex steroid formula probably works best when using plasma. The use of tissue introduces greater variability. While mucus is easily obtained the absolute concentrations are difficult to determine. Of all the biomarkers which can be used probably the vitellogenin and ZRP dot blots are properly the most accurate and user-friendly Recent Further Developments for BFT Hot Blots Fig ZrP-Dot Blot with pab. Plasma was diluted 1/5 and 1/1 with,1m Tris/HCl. Plasma volume was 1µl. 1st Ab-dilution was 1/3, 2nd Ab-dilution 1/15 In recent months further developments have been made for the dot blot and figure 2.14 shows that with further dilution and adjustment of the antibody concentrations and all nothing response can be obtained. The background levels can thus be adjusted such that only female fish are identified. The results for a series of fish from Mazarron are shown in figure At the moment we have refrained from a detailed analysis as done for vitellogenin concentrations as the dot blot test is a subjective measurement and not an empirical one. Its accuracy however is probably an order of magnitude higher. 34

35 A B Nr. Sex Dot- Blot Vtg [mg/ml Plasma] Dot- Blot Vtg [mg/g Muscle] Nr. Sex Dot- Blot ZrP [mg/ml Plasma] Dot- Blot ZrP [mg/g Muscle] 21 M M M F F F F F F M M M Std M M M F F F F F F M M M.18.8 Std. -.1 Fig Results of Hot-Blots for plasma and muscle from Mazarron fish; A is for Vtg measurements and B is for ZRP Conclusions and recommendations: From the data analysis shown above for the different methodologies of sex determination a multiple approach is probably the best however this must be adapted to the problems of sampling either plasma muscle or mucus. The use of the hot blot will under most circumstances gives a positive or negative response thereby making the decision male or female easier to make. In experienced hands this is a very powerful tool for sex determination. As with other methods mentioned here these have been involved over a period of four years and we feel that in skilled hands the opportunity to sex fish is high. We also feel that we have enough material and experiments to offer this service to the industry as a whole. Further development beyond this stage with even more innovation will require substantial funding or joint cooperation between various partners who can each offer expertise and all facilities for testing BEHAVIOURAL STUDIES ON BFT SPAWNING Behavioural studies on BFT spawning using Data-Loggers Introduction When studying the behaviour of captive BFT the most important criteria are to determine when the fish will spawn and in a cage situation when they have released eggs. Normally constant surveillance is time consuming and expensive. It requires considerable ge technological expenditure and on an open sea cage the provision of power sources for video work both on and in the water are limited. Due to the development of data loggers it has been possible to implant recording devices on fish to determine both temperature and depth changes. The rationale behind these measurements are that spawning will usually take place at temperatures above 25 C and an accurate estimation of the length of time spent above 25 C is required therefore the need to continually assess water temperature during the spawning season June / July. The use of external loggers would be adequate for this type of measurement but internal implants could give more information of daily digestion rates and should be less susceptible to loss. Secondly it is known from previous studies in Japan that Females approach the surface of the water at night during spawning bouts so the vertical position in the cage could also gives clues to the spawning behaviour 35

36 of the fish. The present study therefore included both internal and external implantation throughout 23, 24 and 25 using different experimental techniques with varying degrees of success Methodology Internal Application Data loggers: Basics of the data loggers are a max battery life of 2 years, accuracy.1 C and.4% depth accuracy Max 5 m; 1869 double measurement points Data Loggers 23 In 23 a total of four data loggers were implanted successfully within the body cavity. However to catch fish the net was raised and the fish caught by divers, place into a stretcher and brought to the ship deck (fig. 2.16) During this procedure at least two fish died before they could be implanted. Of the implanted fish only the first (ZZ49) survived over 24 hours and was assumed to be present in 25. The other three fish were found dead either in the afternoon after implantation or 24 hours later. No further attempts were made to implant data loggers for this season (23). Fig.2.16 Mazarron cages and caught fish being transferred to deck for surgical implantation Fig 2.17 Insertion position of the of Internal Data loggers and dimension of an individual data logger 36

37 Fig Collage of implanting techniques used involving opening the body cavity and suturing this after insertion of the data logger. Fig Example of programming of internal data logger for fish ZZ49 in 23 Data Loggers 24 In 24 attempts were made to catch fish in cage 3 initially using baited hooks this led to the complication that some very large fish broke the line on numerous occasions. Moving to cage 2 finally fish were removed and in fish ZZ59 with a weight of approximately 65 Kg a data logger (594) was implanted with the following attributes: This fish was opened ventrally and the Data logger inserted using a trochar, afterwards the data logger was suture into position and an external tag placed which was however not fully implanted. The measurement sequence was as follows: FISH ZZ59 Starting time: Feb 1th at 1: (but taken first measurement 2 hours later) Interval period were: 37

38 1: every 15 min (96 data) 24hrs 2: every 12 hours (6 data) 3 Days 3: every 6 min (96 data) 96hrs 4: every 12 hours (3 data) 15 Days 5:every 12 hours (4) 2 days. Measurement interval sequence was since starting time: = Total 496 Days Measurements. On the next day no sign of the external tag was present and the fish could not be identified External Application Data Logger 24 Finally an external data logger was improvised as shown below (fig. 2.19) Fig Prototpye data logger for external implantation using Jab-Stick This was implanted under-water using the jab-stick technique developed in the previous season and used by an experienced diver to implant on the flank of the fish. The results of this prototype experiment were encouraging. The implant remained patent from the until the Then detached and spent some time at the bottom before coming to the surface and being recovered on the The overall results are shown overleaf (Fig. 2.2) for temperature and depth recordings. 38

39 Temp Temperature( C) 46, 45, 44, 43, 42, 41, 4, 39, 38, 37, 36, 35, 34, 33, 32, 31, 3, 29, 28, 27, 26, 25, 24, 23, 22, 21, 2, 19, 18, 17, 16, 15, 14, 13, 12, 1.3 : DA TA LOGGER 591 Implanted free : 1.5 : Measurement time: : Depth, 1, 2, 3, 4, 5, 6, 7, 8, 9, 1, 11, 12, 13, 14, 15, 16, 17, 18, 19, 2, 21, 22, 23, 24, 25, 26, 27, 28, 29, 3, 31, 32, 33, Depth(m) Fig. 2.2 Overall results of data logger

40 Fig 2.21 Individual results Initial 24hrs Depth Temp After 1month After 2 months months 4

41 Fig 2.22 Details of recordings of data logger on and off the fish Tag on surface Depth Tag on Fish Tag on cage bottom Temp Depth Sea surface temperature changes in May during a 24hr measuring period Temp 41

42 After the initial tagging (Fig. 2.21) it is clear that the fish prefer deeper water and in the first 24hrs recording period very few migrations are made to the surface. The water temperature remains at 15 C throughout the measuring period shown in the first 2 months. What can be seen is that with time the fish tend to spend more time nearer the surface during the hours of darkness with a return to deeper water at around 2meteres at the bottom of the cage. This appears to take place at approximately 6: hours and excursions to the surface again at 21: hours. Exkursions are made towards the surface but mostly only to a depth of 2 metres. In figure 2.22 the fate of the tag is described in detail. For some unknown reason it became detached from the fish. Probably the nylon monofilament broke but although the tag was positively buoyant it remained at 2m on the bottom of the cage for 2 days. Afterwards it continued to the sea bed (3m) where it remained for 1 days before finally surfacing where it recorded sea-surface temperatures over a continuous period (Shown in the lower half of the figure). The surface temperatures show a clear pattern of heating and cooling in the middle of May with low temperatures of 15 C at night compared to a maximum of 27 C between 13: 14:. Data loggers 25 In February 25 a new attempt was made to implant Tuna with external data loggers. The newly designed system is shown below (Fig.2.23) and consisted of a polypropylene body with a nylon monofilament leader attached to a Floy Arrow head. The arrowhead was large enough to accommodate a PIT tag. The propylene body is made up of a screwtop through which the Star-Oddi DST milli - TD logger Small Temperature and Depth Logger could be inserted. The leader consisted of a slider which held the monofilament link together followed by a medical silicon tube together with a penetration marker (8cm) and then colour coded tubing (see figure below). Fig The pictures show the Data logger from Fish ZZ months after implantation. Note the algal growth on the attachment tether and also a mussel spat growing on the top of the Data logger. In total three fish from Cage 2 were implanted under-water using a spear-gun technique on the On the one of the tags had become detached. The other 2 tags remained in place until the when the fish were recaptured at slaughtering. The tags were set with the following measurement settings for sampling temperature and depth. Measurement start time : :: Measurement settings: [dd:hh:mm:ss] x number Start delay : :15:57: 1. interval period : :15: x interval period : 12:: x 6 3. interval period : :3: x interval period : :12: x interval period : :6: x interval period : :12: x 5 7. interval period : ::1 x 6 Measurement interval sequence: The results of the measurements are detailed in the following diagrams (fig 2.24 and 2.25): 42

43 Fig The data shown in the figure above summarizes the total recoding of depth and pressure in the two tagged fish, ZZ 22 a male fish which was tagged and induced, ZZ 211 a female control fish. The dark arrows indicate an event which seems to have taken place displacing the temperature axis. Recordings were also made at periods over 24 hours every 15 minutes. Fig 2.25 The data showed in the above picture illustrates a direct comparison between the treated male and the untreated female control. The dotted red line indicates the maximum/min depth attained by the fish. The dark bar indicates periods of darkness. 43

44 From the recordings is clear that both fish follow similar patterns of activity although the female fish remains of the lower depth than the male fish. There appears to be activity peaks as the sun rises as can be seen in the temperature recording shown in red. At sunset there appears to be a trend in both fish to approach the surface. This may be indicative of spawning behaviour or the precursors of such activities. A further detailed analysis has been made whereby the data for the first four months was analysed using a 12 hour measuring period and the mean values taken (Fig. 2.26). These indicated that at the beginning of the measurement period in February water temperatures were relatively cold at around 4 C. The Fish ZZ 211 spent a relatively large amount of time at the bottom of the cage with infrequent visits to the surface. In March surface temperatures began to warm up and the fish appear to spend more time in the zone between 16 and 8 m. At night time there was a tendency to be higher in the water column, i.e. around 8 m, than during the daytime where the majority of the time was spent below 12 to 16 m. In April and May water temperature rose but parallel to the depth axis of the rate of approximate 3 per month. During the period May less than 5% of the total time was spent at the surface. It must be remembered however that these measurements were made at 12 noon and 12 midnight and not continuously. A period of 24 hours was measured once a month where the recording frequency was every 15 minutes. As a spawning season approached the data logger was switched to a higher measurement frequency, depth and temperature being recorded every six minutes. A similar picture to the one above was also obtained for Fish ZZ 22. Since both fish ring the cage number two this is not surprising that there were individual differences in the preferred depth of the fish possibly indicating the schooling behaviour already observed in cage fish. Fig 2.26 This figure shows a more detailed examination of the 12 hour recording data for the months from February and May 25 in the female Fish ZZ 211. Temperature with depth data are shown by the linked points, no standard deviations are shown. The horizontal bar diagram shows the percentage of occurrence at a given depth classes between zero to 4 m and 16 to 2 m. From the more intensive measurement periods a clearer picture is given of the water conditions and the behaviour of the fish themselves. During the middle of June a thermocline was formed within the water 44

45 column. This can clearly be seen in the figure with the water temperature proximate 2 to 3 higher at the surface (4 m) compared with a depth of 16 to 2 m. The temperature profiles in both fish, as expected are almost identical. In June both fish only approach the surface during periods of darkness, during the daytime they spend more time below a depth of 8 m. On the whole in June and July the majority of time at night is spent above the 8 m mark. It is clear from measurements in both fish that the preferred spawning temperature of 25 C is reached only in July, and then at a depth above 12 m. These points should be taken into consideration when considering the spawning window of the tuna. Through the use of the implanted dataloggers it has been possible to track spawning behaviour and in general observe the behaviour of individual fishes. Data Logger#59- ZZ 211 Female 6 min Measurements June July 25 Fig This figure shows the recordings made at 6 minute intervals for both ZZ 211 female and ZZ 22 Male fish for the months of June and July. Temperature with depth data are shown by the linked points, no standard deviations are shown. The horizontal bar diagram shows the percentage of occurrence at a given depth classes between zero to 4 m and 16 to 2 m. The dotted line indicates the suggested 25 C preferred spawning temperature. From the handling point of view it is clear that the baited hook will only succeed for catching a few fish. The surgical procedures for Data Logger implantation need to speed up to decrease the time out of water to less than 4 minutes. The skin of the tuna is extremely tough and dedicated sutures and needles must be further exploited to speed up the surgery on deck. Of the sampling season in 25 thought should be given to trying out further handling techniques. During the third reporting period and new series of data loggers with improved attachment systems together with a capsule for the data-logger itself were deployed on three fish. Two of the deployed data loggers remained in place for a period of five months from February until July 25. One data-logger detached 24 hours post deployment. The data loggers were implanted using a harpoon system rather than a jab stick as previous attempts. All the deployed data loggers one was attached to a female the other to are male fish. The data-loggers had been individually programmed such that the battery and the number of data acquisition points were said that during the period from February until May recordings were made at a relatively lower rate. In June the chip was switched on to record both depth and temperature a period of six minutes 45

46 Conclusions It could be confirmed that the use of data-loggers provide an excellent means of monitoring behaviour of the BFT during the spawning periods. The information obtained on temperature changes and the presence of a thermocline would not have been evident if just single one point measurements had been taken. The added advantage of being able to monitor the position of the Fish in the water column at regular intervals is also advantageous. Since the fish carries the data-logger itself and the recording periods can be programmed, this does away with expensive in situ recording systems which may be stolen or removed from a cage structure. The data-logger has also the added advantage that it records the actual temperature regime of the fish itself and not the surrounding water. This is very important if one is looking at spawning at particular temperature regimes Studies of broodstock swimming behavior by echosounder system Methodology and study material In order to study the vertical movement of the BFT broodstock in the floating cage, and eventually detect any presumable reproductive behavior, an echosounder system was used. The system was composed by a 1x2 cm 38 khz ceramic transducer attached to a platform floating in the center of the cage and connected to a transceiver (Fig. 2.28). Data was stored in a laptop computer (Fig. 2.29). Energy for all the system was provided by two batteries. All these devices were located into an hermetic stainless steel box fixed on the external ring of the cage. Fig Echosounder systems used for study the BFT swimming behaviour. 46

47 Fig Diagram showing the transfer of echosunder data In the morning, the system was daily turned off and data was downloaded (fig. 2.3). The monitoring was made between February and April 25 and June July 26. Tunas were fed with raw fish once a day; around 1: a.m. Binary data provided by the transceiver in the daily files was observed with the Sonar Data Echoview software. Fig. 2.3 Daily downloading of the echosounder data obtained in the previous day Results and Conclusions Fig shows a typical diagram of the echosound observed. 47

48 TIME (HH:MM:SS) 1:6: 1:6:15 1:6:3 SEA SURFACE 1:6:45 1:7: 1:7:15 1:7:3 1:7:45 1:8: 1:8:15 1:8:3 1:8:45 1:9: 1:9:15 1:9:3 5 TUNAS DEPTH (m) BOTTOM OF THE CAGE SEA BOTTOM 3 Fig Typical echosound diagram. The horizontal axis corresponds to time and vertical axis corresponds to depth (meters). Two continuous bands in high density color appear. The one at 3 m corresponds to the sea floor and that at 2 m in the middle of the echogram to the bottom of the net. Single or groups of tuna fish showed high density acoustic stains. During the monitoring period BFT seems to activate at sunrise before the feeding time, with so frequent visits to the surface (Fig. 2.32) TIME (HH:MM:SS) 7:: 7::15 7::3 7::45 7:1: 7:1:15 7:1:3 7:1:45 7:2: 7:2:15 7:2:3 7:2:45 7:3: 7:3:15 7:3:3 5 DEPTH (m) Fig Diagram showing BFT specimens near the water surface. This activity seems to be reduced at sunset, resting the BFT at the deeper half of the cage at night with occasional visits to the surface. Fig represents a group of tunas passing through acoustic beam. But, if 48

49 acoustic beam (transducer) is in the center of the cage, and tunas are swimming around the cage, it probably represents only a part of entire group of tunas that in regular period pass below transducer. TIME (HH:MM:SS) 2:: 2::15 2::3 2::45 2:1: 2:1:15 2:1:3 2:1:45 2:2: 2:2:15 2:2:3 2:2:45 2:3: 2:3:15 2:3:3 5 DEPTH (m) Fig Echogram of BFT located in a deeper position inside the cage, concentrated into schools. Fig shows BFT which are swimming in schools, but less compact. The shape of this group (school) is more extended horizontally. This pattern probably could be explained by the fact that different schools swim at different depths of the cage, passing below the transducer at more o less equal time periods. TIME (HH:MM:SS) 7:: 7::15 7::3 7::45 7:1: 7:1:15 7:1:3 7:1:45 7:2: 7:2:15 7:2:3 7:2:45 7:3: 7:3:15 7:3:3 5 DEPTH (m) Fig Echogram showing BFT swimming in a less compact shool. Some of the echograms obtained were difficult to be interpretated. The strange images obtained were associated to water turbidity, bad sea conditions, and influence of the BFT feeding activity and/or interferences by navy ships. 49

50 Turbidity conditions In Figure 2.35, blue color represents weak echoes. If echosounder is set to minimum value to -7 or -8 db, this "blue echo" is probably originating from plankton and/or detritus (particular organic matter). Normally, plankton has diurnal (daily) vertical migrations from bottom to the surface, and consequently this blue zone is changing its depth. TIME (HH:MM:SS) 7:: 7::15 7::3 7::45 7:1: 7:1:15 7:1:3 7:1:45 7:2: 7:2:15 7:2:3 7:2:45 7:3: 7:3:15 7:3:3 5 DEPTH (m) Fig Echograms obtained under water turbidity conditions. Bad sea conditions The reasons for the abnormal echogram showed in Fig are probably movements and/or a not perfectly horizontal position of the transducer under rough sea conditions. Apparently on the echograms, sea surface is always a straight line, but movements of transducer are affecting bottom echoes. TIME (HH:MM:SS) 7:: 7::15 7::3 7::45 7:1: 7:1:15 7:1:3 7:1:45 7:2: 7:2:15 7:2:3 7:2:45 7:3: 7:3:15 7:3:3 5 DEPTH (m) Fig Strange echograms eventually related with bad sea conditions. 5

51 Influence of the BFT feeding activity At these echograms (Fig. 2.37), it can be distinguished echoes of individual BFT, where tunas are not grouped (in schools). TIME (HH:MM:SS) 7:: 7::15 7::3 7::45 7:1: 7:1:15 7:1:3 7:1:45 7:2: 7:2:15 7:2:3 7:2:45 7:3: 7:3:15 7:3:3 5 DEPTH (m) Fig Echogram with dispersed BFT fish associated with feeding activities Interferences from Navy ships The usual presence in the area of ships from the Spanish Navy may have produced interferences because of its own sonar systems (Fig. 2.38): TIME (HH:MM:SS) 7:: 7::15 7::3 7::45 7:1: 7:1:15 7:1:3 7:1:45 7:2: 7:2:15 7:2:3 7:2:45 7:3: 7:3:15 7:3:3 5 DEPTH (m) Fig Strange echogram associated with interferences made by navy ships. In conclusions, the echosounding system may be helpful for monitoring vertical movements of tuna broodstock along the day, but it will be necessary to enhance the reception of echosound, thus discriminating the noise coming from the nets, especially at bad sea conditions. 51

52 It must be expressed our gratitude to Dr. Vjekoslav Ticina (IOF, Croatia) who collaborated in these studies. 2.3 SET UP SUITABLE ANAESTHESIA AND SEDATION PROCEDURES FOR BFT Methodology and study material Efficiency of common and new fish anaesthetics were tested in a small number of BFT specimens and other species with similar size supplied by local fishermen. Several application techniques for the anaesthetics (bath immersion, injection) were employed, as well as suitable dosage tried to be determined Results Clove oil as a safe and effective anesthetic has been tested in several marine reared finfish species (Dicentrarchus labrax, Sparus aurata, Dentex dentex, Pagellus acarne, Seriola dumerili). With respect of Seriola dumerili, the effectiveness of this novel anesthetic was proved on fish from few grams up to 4 kg. Optimal dose is usually 4 ppm. Anesthesia and recovery time is normally only a few minutes. Beside the low cost of this product, on of its main advantage is that the fish anaesthetized can be consumed right after treated, since clove oil is a natural spice and no withdrawal period is required. This is quite important for the commercialization of a so high value fish such as BFT. An attempt to the use of clove oil with BFT was made on February 23. Two BFT (mean body weight 7 kg) caught from the experimental cage by hooking were placed into a plastic container of 1 l with 2-8 ppm clove oil. A mask fitted to the BFT mouth and connected to a water pump; provide a constant flow of the dissolved anesthetic through its gill. Fish get into anesthesia after 1-15 min of immersion on the bath (much longer than in other fish species) and recover in about 2 minutes. Results were unsatisfactory, but it was considered not related with the effectiveness of the product itself but with an unsuitable administration procedure. Since previously to start the project, clove oil was used on same size BFT and resulted quite effective. Anyway, to avoid the risk of loosing valuable broodstock specimens, bath anesthesia was eventually discarded. Regarding the use of injected anesthetics, a thoroughly survey of existing scientific literature on such topics, not just in fish but in other animals (mammals, birds, reptiles and amphibians) was made. In the early summer 25, some experiments were conducted in order to check the efficacy of some anesthesia procedures for large pelagic fish, such as BFT and amberjack (Seriola dumerili). The use of injected anesthetic was the methods chosen since the difficulties of handling such a large fish (from 2 to 2 kg) make unsuitable others usually employed in aquaculture operations like bath anesthetic, In fact, the trials made on such technique during the first reporting period resulted in the death of tunas because of acute stress. For the application methods, the use of jab stick specially conceived for the restraint and immobilization of wild animals and zoo applications was tested (Fig. 2.39; Dan-inject Aps, Denmark). This is an automatically discharging jabstick which injects 1-1 ml in one second or less in contact. It features an easily cocked trigger device that releases the injection on contact with the animal. The safety lock ensures safety for personnel. The jabstick is 1 meter in length with a removable extension handle that increase the effective working distance to 2 meters. It utilizes nylon multi-use luerlock syringes. Fig Discharging jabstick from Dan-Inject employed for anaesthesia application in BFT. 52

53 With respect of the injection needles, the main problems were that when fish are punched they suddenly make a body contraction to escape. And then, the needle usually is bent or broken. It was used different sizes stainless-steel needles (from 2.x 15mm to 3.x4 mm). The 3.x25mm needle resulted the most effective one. Regarding the drug anesthetics, it was used Ketamine hydrochloride and Detomidine and its antagonist atipamezol (Domosedan and Antisedan, Pfizer) Ketamine is a cyclohexamine that has been utilized in several Pacific Teleosts and elasmobranch species (Stoskopf, 1986; Williams et al, 1988, Graham and Iwama, 199; Fleming et al., 22) as well as in some tunid species (Meghan and Block, 1998; Williams et al., 24) when long-lasting anesthesia is required (up to 48 h) without blocking the ventilatory rhythm system, and because it has been demonstrated that it has a large safety margin before becoming lethal. Detomidine is an alpha2-adrenegic agonists, a non narcotic sedatives providing good analgesia and muscle relaxation when used in horses and cattle. This drug are extremely useful on their own but find their most common use in combination with anesthetic agents like ketamine or with the opioids used in wildlife capture (carfentanil abd ethorpine) where they temper the bad side effects notable with these other compounds. Detomidine effect is reversal with atipamezole. During the trials it was tested the effect of Detomidine (1mg/ml), Ketamine (1g/1 ml) alone or in combination, and Atipamezol (5 mg/ml). Amberjack size were 3-4 kg body weight, while the three BFT used as experimental animals (from the broodstock) were around 1 kg. The dosages employed intramuscularly were 6,12 & 2 mcg/kg Detomidine, 2-3 mg/kg Ketamine and.3-.6 mg/kg atipamezol. Because of the actual concentration of the commercial presentation, it was required to make several injections (up to 1 ml) to reach the optimal doses Conclusions Even though the anesthetic cause sedative and immobilization effect on the treated fish, it required at least 3 min to produce restraint. Moreover, recoveries of the fish were not satisfactory and some of them died 24 hours later. It was concluded that neither the application method nor the drug anesthetic was enough effective for the BFT anesthesia. Further experiments should be focused on the use of dart syringes, having a gas or air chamber, propelled by speargun. Since use of jabstick have some concerns to ascertain the drugs absorption by the muscle because sometimes the neddle was bent. Furthermore, it should be tested other more potent anesthetic, like opioids, to make fish become anesthetized in shorter time. Unfortunately, no more BFT were available to carry out these studies. 2.4 DESIGNING TRANSPORTATION SYSTEMS FOR BFT AT OFFSHORE AND TO IN-LAND FACILITIES Catching methods Methodology and study material The setting up of a safe and effective methodology for catching brood BFT specimens from the experimental cage was attempted. Two main approaches were employed: chase and stretcher and hook and line. For the first one (Fig. 2.4), some fish were herd to a transfer cage attached to the experimental cage. This consisted on a square net (15x9x12 m), having a window or door that can be connected to another one at the circular cage and opened both for doing the fish transfer. Once there, cage bottom is lifted to gather and concentrate fish into a more reduced area (1-2 m deep) and chased until being trapped and immobilized by some divers into a stretcher made by soft plastic vinyl. There, BFT are put belly up and close eyes to make fish calm. Stretcher roped to the ship crane is lifted aboard and fish placed over a foam mat. If it s required to keep fish alive, a mask designed for use on big size BFT is employed. This consisted on three hoses: one connected to a water pump, another to be placed on the fish mouth and the other with a valve that can be opened to adjust the water flow passing trough the gills. 53

54 Fig. 2.4 Use of transfer cage for chasing the BFT specimens and transferring into the ship s deck by stretcher. The hook and line methods (Fig. 2.41) consisted on a barbless hook baited with raw fish. Once the fish bite the hook, line is pulled to bring it to the cage border. Then, the fish are immediately led into the stretcher, hook removed and lifted aboard. 54

55 Fig Use of hooks and line methodology for catching BFT specimens Results and conclusions Hook and line method was proven more safe for the fish, since less mortalities happened than with the use of chase and stretcher. Main disadvantage is that some times the fish release the hook causing them wounds. Furthermore, it is almost impossible to take up many fish by this method, because all fish do not have always good appetite and eat the baited hook. It has been also observed that when some fish unhook, the remaining one at the same cage soon avoid it. When handling the fish for tagging and it was required to keep fish alive, it were sometimes employed a special mouth designed for BFT that provided a constant flow of seawater through the gills to avoid it suffocation (Fig. 2.42). This consisted on three hoses: one connected to a water pump, another to be placed on the fish mouth and the other with a valve that can be opened to adjust the water flow passing trough the gills. Fig Use of water flow mask to keep BFT alive. By other hand, those fish captured at the transfer cage by chasing have resulted in higher cortisol and cathecolamines levels (see results in the next chapter, indicating that fish received an acute stress. As conclusion, hook and line method agreed to be the most suitable one for handling BFT for sampling purpose on a few specimens. One of the major problems is to design a method of removing the fish from the cages were the least possible stress. Chosen of safe and effective methods for handling operation have resulted sometimes difficult. Then, use one or the other already tested was depending upon the aim of it (sampling dead animals, tagging alive fish, and so). It would appear at the present time that a baited hook and lifting the animal on a stretcher are the method of choice for the future because some physiological parameters or even the survival possibilities of BFT were shown strongly influenced by the methods employed; new approaches about it might be considered in the near future Transportation methodologies Transportation trials for BFT and other big pelagic fish species were approached. In that sense, specialized staff from the large aquarium of Valencia (Spain), L Oceanografic, leaded by Mr. Pietro Pechioni (Head of Pelagic and Ocean Section) collaborated with us. They were interested in capturing and transporting such fish from the Mazarron area to their facilities (35 km far away). That is why covered all the expenses 55

56 regarding renting fishing boats and crew, transportation (equipment and track), and their own personnel and accommodation. IEO provided with personnel and facilities for the acclimatizing period before transportation. Target fish species were some Scombrids (Thunnus thynnus, Sarda sarda, Euthynnus alletteratus, Scomber japonicus, Scomber scomber, Trachurus trachurus) and other large size fish like greater amberjack (Seriola dumerili) swordfish (Xiphias gladius), dolphin fish (Coryphaena hippurus) and sunfish (Mola mola). Fish were captured from April to May on a traditional set net ( Almadraba ) settled in La Azohia, Cartagena, SE Spain (see fig. 2.43). Either plastic bags or a special vinyl fish nets were used to transfer fish to a tank installed on board a boat. Then, they were brought to the shore and transported by track IEO facilities. Fish were injected with Dexametasona (anti-stress therapy) and an antibiotic (Ceftazidima) to reduce the risk of losses. Fig Fish capture at the Almadraba trap and transportation to IEO inland facilities. After an acclimatization period of about one week in a 4 cubic meter concrete tank, they were transported to the L Oceanografic facilities by lorry (Fig. 2.44). Fish were introduced into 35 l cylindrical tanks, supplied by a 6 l/h pump with carbon filtered recycled water. Dissolved Oxygen concentration was maintained between 15 and 18% and water temperature kept constant at 16ºC. Fig Transportation of pelagic fish from IEO to L Oceanografic Aquarium in Valencia (Spain). After a 4 hours trip by road to L Oceanografic, fish were transferred to tanks to the quarantine area. There, fish were treated with antibiotics and disinfectans every 48 hours and fed with Sardinella pilchardus, 56

57 Sardinella aurita, Sprattus sparattus, Engraulis encrasicolus and Loligo sp. Few weeks later, fish were brought to the 69 cubic meters exhibition tank measuring 33x95x5.5 m (Fig. 2.45). Fig Pelagic fish species transported from Mazarron were under quarantine until transferred to the exhibition tank. As result of these trials, 14 bonitos (Sarda sarda) were caught. 78 survived the transportation and 59 the quarantine period. Thus, a 6% survival rate was obtained, which is considered quite positive since this species is very sensitive to handling. Several specimens from others species were also successfully stocked on the exhibition tank (Fig. 2.46). Major difficulties were encountered during the transportation of swordfish. A B C D 57

58 E F Fig Pelagic species which were successfully transported to L Oceanografic. A) bonito. Sarda sarda; B) little tunny, Euthynnus alletteratus; C) mackerel, Trachurus spp.; D) dolphin fish, Coryphaena hippurus; E) sunfish, Mola mola; and F) swordfish, Xiphias gladius. From the end of August to October, several shipments were made to capture young BFT-of-the-year and dolphin fish. Unfortunately, it was not possible to catch BFT. On the other hand, some small dolphin fish were caught and successfully transported to the aquarium in Valencia. The lack of capturing young BFT has prevented to carry out further transportation trials and develop suitable methodologies adapted to its specific characteristics. Nevertheless, the ones successfully accomplished in other tuna-likes species (bonito, and other Scombrids) may be beneficial for the above. In spites that it were tried to catch wild small BFT by hooking fishing, to carry out transportation trials, there was no success. Then, no activities on the matter were conducted. 2.5 DEVELOPMENT OF A METHOD FOR COLLECTION OF EGGS RELEASED NATURALLY IN THE CAGES Results in 23 During the first year and in order to collect any eggs spawned by the BFT breeders, some collectors were installed on the experimental cages. These consisted on plankton net collectors, usually employed for coelenterate larvaes. Mesh size of the nets was 1 mm and 5µm at the end of the pvc tube for collection. Its mouth having 1m in diameter (see fig. 2.47). At the end of the net an open bottle was placed to keep any eggs available. The collector was oriented against the main water current and roped to the cage anchoring cable and to the net of the cage. From June until mid July, the collectors were daily inspected in order to find any eggs. Net was rinsed by seawater and the content of inside the bottled fixed with buffer formalin. Unfortunately eggs found did not seem to come from BFT origin. Blind sheet.5 m length Tube 1 m Fig Design of BFT egg collector attached to the cage net. Net Mesh size5 PVC tube 16 mm Net 2 m length Mesh size 1mm Tube 1 m 58

59 Results in 24 In order to collect any eggs spawned by the caged BFT breeders on the second year, former and new approaches were accomplished (Figs and 2.49). As far as egg collection is concerned, a plankton net collector (5µm mesh size) with a squared mouth of 5 cm was used to sample the seawater surface inside the cage. Another plankton net of a circular mouth of 1 m was towed near the surface in order to screen the water around the cages. Besides that, a net of a fine mesh size was passed throughout the cage surface. Moreover, a black plastic sheet of about 2 m high was placed all around the cage rings (Fig. 2.49). Furthermore a new egg collector device of 12 m total length and 6 m mouth were made with a decreasing mesh size when reaching the blind end (Fig. 2.5 and 2.51). These were roped to the net of the cage. EGG COLLECTION DEVICES Tube 1 m Blind sheet.5 m length Net 2 m length Mesh size 1mm Net Mesh size5µm Tube 1 m PVC tube 16 mm Fig Devices employed in the collection of BFT eggs (plankton net collectors) in 24. Fig Black plastic sheets fitted to the cage to prevent the escape and collect the BFT eggs. 59

60 6 m 6 m Ropes Sea level FRONT VIEW BFT EGG COLLECTOR DEVICES Sea level LATERAL VIEW Ropes Hole for rinsing/flushing out the blind end Net mesh size 1mm Floats line Sinkers line 6 m Collector 12 m in length with net mesh decreasing in size AERIAL VIEW Fig. 2.5 Design of the large device employed in the collection of BFT eggs in 24. Fig Underwater view of the large device employed in the collection of BFT eggs in 24. Like in the previous year, some observers were daily, from June 15 th to July 15 th, at the experimental cage facilities two hours before sunset and one hour after. There, they observed the broodstock behaviour and tried to find out any putative spawning one (quick swim, fish chase, jumping out of the water). Surface water temperature was recorded and the sea condition (wind and waves regime). Then, water was sampled by the plankton net collector and the net seine, and checked the egg collector device looking for the occurrence of any eggs (Fig. 2.52). If any, net was rinsed with seawater and eggs put inside a bottle with buffered formalin. Eggs were then inspected and measured under a binocular microscope to determine their possible origin. 6

61 Fig Daily screening of the water surface inside BFT cages at sunset for the collection of any putative eggs by the use of fine mesh nets and plankton collectors. As result, several eggs were collected, but none of them appeared to have a BFT origin (Fig. 2.54). The large eggs collector resulted unsuitable and easily damaged by the high speed current. No. Eggs collected Date , , , ,5 T (ºC) No. eggs T(ºC) Fig 2.53 Number of fish eggs collected and seawater temperature at the experimental cage site in summer /6 25/6 27/6 3/7 5/7 6/7 9/7 28/6 28/6 29/6 7/7 12/7 1/7 13/7 29/6 2/7 14/7 15/7 Fig Eggs collected in 24 belonging to other fish species, not BFT. 61

62 Major difficulty was the lack of BFT spawned eggs collected. Beside the non.-occurrence of spawning large batches of eggs, another possibly reasons for that might be either strong seawater currents which spread out the eggs very quickly outside the cages area, or a putative egg depredation behaviour made by the opportunistic finfish species living around Results in 25 Along 25 spawning season significant efforts, both in labor and purchasing materials, were deployed in order to improve the capability to collect any putative spawn released naturally or hormonally induced by the BFT breeders. A new approach for the eggs collection inside the rearing cage which could be eventually spawned by the BFT breeders was tackled this year. This consisted on the utilization of a barrier installed around the cage perimeter which may prevent the eggs to move out the cage enclosure. A similar technique is usually employed by Japanese BFT researchers. One of the main concerns about it was that the curtain could damage the cage structure or affect its mooring by exerting a strong resistance to the seawater flow, particularly when high speed currents would happened. Two kinds of curtains were designed. The first one was made with a mixture of vinyl plastic sheets combined with fine mesh nylon nets (355 and 5 µm).the curtain has islets every 3 cm at the upper and down side to rope it to the net of the cage on the inside. Along the sheet of about 5 m length and 4 m height, it was installed 4 "copos" to collect the eggs, a kind of plankton net conical in shape, each one separate to the other by 9 m (Fig. 2.55) The second one differs slightly than the other only in the fact that instead of using plastic sheets, it was used a combination of nets of different mesh size. (Fig. 2.56) It was expected that by using a blind or a fine mesh net, this would probably facilitate the slide of the eggs towards the copo. Thereafter, eggs collected at the copo could be easily obtained by removing a PVC tube fitted to the top end of the net with that has a 5-µm mesh on its bottom (see Fig. 2.57) The plastic curtain covered more than 2/3 of the cage perimeter and it was oriented to the north-west (main wind), leaving free the south-west area to facilitate the water flow It was made an assay to check the efficacy of the plastic curtain to retain eggs released inside the cage. Some trials were conceived to asses the way which fish eggs might be displaced inside the cages under a regular water current regime. Gilthead sea bream eggs (similar in size of BFT ones) were released at different depth by divers, and then checked its spread, buoyancy capacity and speed to move out from the cage area. It was used 1 million of sea bream eggs, which were unfertilized and 4. floating eggs, thus fertilized. Eggs were released at about 2 m depth on the center of the cage. The current was from moderate to slightly strong. Once the eggs were released, their movement was tracked. The sinking eggs (unfertilized) were chased and quickly eaten by small opportunistic fish. In spite that there were less abundant than in previous years (about two hundreds) since from early spring a fine meshes size net was placed inside the cage to avoid the entrance of big fish predators. The fertilized ones were dispersed vertically and the current made it reach the cage peripheral in 2 min, under the 2 m depth, down to the plastic barrier placed around the cage for collecting eggs. Half an hour later the surface of the curtain was screened with a net, but it was not found any eggs at all. It was concluded that the eggs were not positive buoyancy but neutral or slightly negative (because the effect of the sea current) and pass through the cage deeper than expected. According with the former results, it was decided to place a completely blind black plastic curtain from the surface to 2 meters and then, the net collector up to 4 m more in deep. Thus, having a barrier of 6 m in total from the water surface. 62

63 7 m 4 m 1 m 4 m 1 m 4 m 1 m 4 m 1 m 4 m 1 m 4 m 1 m 4 m 1 m 4 m 7 m 4 m 53 m PVC Fine mesh Large mesh 25 m Fig Egg collection curtain made of different mesh size nets and PVC sheets 63

64 45 m 4 m 5 m 5 m 5 m 5 m 5 m 5 m 5 m 5 m 5 m 25 m Fig Egg collection curtain made of different mesh size nets 64

65 Islets PVC 1 m Large mesh 4 m 2 m Fig Details of the Copos at the egg collection curtain. 65

66 Apart from the curtains, it was also followed the same approach than in previous years for collecting eggs. That is: To use of a plankton net collector (5µm mesh size) with a squared mouth of 5 cm to screen the seawater surface inside the cage. To tow a fine mesh net seine (over 15 m in length) all across the cage surface. To screen the seawater surface around the cages installation by towing another plankton net collector with a circular mouth of 1 cm. Like in the previous year, from June 15 th to July 15 th some observers were daily at the experimental cage facilities two hours before sunset and one hour after. There, they observed the broodstock behavior and tried to find out any unusual one that could be related with spawning (quick swim, fish chase, jumping out of the water). Besides, surface water temperature and sea conditions (wind and waves regime) were recorded. Thereafter, seawater was sampled by the plankton net collector and the net seine, and checked the egg collector devices looking for the occurrence of any eggs. If any, net was rinsed with seawater and eggs put inside a bottle with buffered formalin. Eggs were then inspected and measured under a binocular microscope to determine their possible origin. Furthermore, the top end buckets of the nine copos at the two curtains installed into the two cages were daily removed by divers and its contents rinsed and put inside bottles. As result, several eggs were collected, but only a few of them appeared to have a BFT origin. Some of these eggs were collected from the curtain installed at cage 1; and appeared to be fertilized. It were brought to the IEO facilities and incubated. Eggs hatched and larvae developed until yolk absorption. Morphological characteristics of the larvae seemed to indicate a possible BFT origin, after consulting expertise BFT fisheries biologists from IEO organization (Alberto Garcia, personal communication). Anyway, it can t be concluded that the new approach of installing net barriers were unsuitable to collect eggs inside cage. Since eventually, either a very few eggs were spawned or could have been predated by small fish living around. Future trials should be conducted to check better its efficacy, avoiding the existence of small eggs predators inside the cage and having a bigger number of BFT breeders that may spawn greater number of eggs. The positive conclusion was that the cage structure and mooring did not suffer any damage because of the installation of egg collection barrier. 2.6 STRESS LEVELS OF CORTISOL, CATECHOLAMINES AND LACTATE IN THE PLASMA OF CAPTIVE FISH Introduction Fishes display a wide variety in their physiological responses to stress. Primary endocrine responses to acute stress include the release of catecholamine and corticosteroid hormones into circulation. After a cascade of neuroendocrine events initiated by perception of the stressor, corticosteroids are released by the interrenal tissue, the adrenal homologue in fish, which is concentrated in the extreme anterior portion of the kidney in teleosts. The primary corticosteroid released during stress in both teleostean and chondrostean fishes are cortisol whereas in elasmobranchs, it is 1 alpha-hydroxycorticosterone. Elevation in circulating cortisol during the first hour after an acute disturbance can increase from relatively low resting levels to between about 2 and more than 1 ng/ml, depending on species. Plasma cortisol is a good indicator of the magnitude of the stress response. In freshly caught BFT values for Adrenaline ranged around 1 nmol/l compared with noradrenaline values around 3 nmol/l (65). In a variety of fish the absolute values vary considerably and in the Plaice very little increase was found even after trawling (65). Since BFT are extremely active only general comparative levels of Catechols can be determined but the maximum thresholds will give us some idea of stress responses and a comparison between caged and wild fish. Cortisol normally increases more slowly but the monitoring of both parameters will enable us to see if these factors effect reproduction in general giving rise to atresia in the gonads. The third variable which will be measured in the plasma and tissue was the lactate concentration. This metabolite is formed during anaerobic respiration which is caused by the lack of oxygen supply to the tissues. This may be caused by increased activity due to swimming and is relatively fast acting in its response in most fish. Again the amounts of lactate and the kinetics of release are dependent on a number of factors but there is a clear correlation with a length of duration of the stress and the level reached Material and Methods 66

67 The three basic criterion for stress used in this study were the levels of Cortisol, the Catecholamines Norepinephrine and Epinephrine, and Lactate in the plasma. The use of these three criteria was chosen to represent chronic and acute stress parameters. Basically cortisol levels may arise within a one to four hour time frame and then remain relatively high, Catecholamine rise relatively rapidly and also lactate levels within 3 minutes. Standard assays were performed using ELISA techniques for Cortisol and HPLC techniques for Catecholamines. Lactate was assayed with a standard optical test. Catecholamine Measurements Adrenaline and Nor-adrenaline Plasma samples were extracted using standard procedure with acid activated aluminium powder, washing an then extraction with perchloric acid. Sample will then be measured with HPLC and electrochemical detection. External standards for Epinephrine and Nor-epinephrin were used together the internal standard DHBA to control extraction efficiency. Cortisol Measurements A standard cortisol ELISA was used: This assay is based on the competition principle and the microtiter plate separation. An unknown amount of antigen present in the sample and a fixed amount of enzyme labelled antigen compete for the binding sites of the antibodies coated onto the wells. After incubation the wells are washed to stop the competition reaction. Having added the TMB substrate solution the concentration of antigen is inversely proportional to the optical density measured. The measured ODs of the standards are used to construct a calibration curve against which the unknown samples are calculated. Plasma and Muscle Lactate Standard biochemical technigues whereby L-lactate in plasma, seurm or muscle can be measured by enzymatic conversion of L-lactate to pyruvate with concomitant conversion of NAD to NADH, the increase in absorbance at 34 nm being proportional to NADH formation. Lactic acid and glycine buffer (ph 9.2) can be obtained from Sigma-Aldrich Co. (St Louis, Missouri, USA) NAD (solid; 2 mg/ml) and lactate dehydrogenase (15 ku/ml; 5 U/ml) from ICN Pharmaceuticals Ltd Results Mazarron 23 On the four different sampling dates from March through till July different methods were used to restrain the fish. a) First sampling March: These fish had been restrained within a confined sampling closure/ transfer net for 24 hours and then lifted out of the water using a transfer stretcher. They were sampled on deck and then replaced in the water. These are the fish numbers ZZ1 to ZZ3 from which blood samples were obtained. On the following day all three fish had died and they were sampled again. Fish number four ZZ4 was also found dead in the cages. b) Second sampling May: These fish ZZ31 36 were caught on baited hooks and immediately removed from the water and sampled. Fish ZZ37 and ZZ38 died due to transferring Fish ZZ38 and ZZ12+ died after they had been tagged the previous day. c) Third sampling June: In this sampling period fish were caught by raising the net enclosure. Fish ZZ 44 however was killed using a harpoon. d) Fourth sampling period July: During this sampling period the net was raised again and the fish caught by entrapment. This led to much chasing of the fish around the cages. During this time the fish receive the highest acute stress. Lactate and Cortisol From Figure 2.58 it is evident that in the first, third and fourth sampling period s considerable stress was involved as shown by the increases in plasma cortisol and also in lactate. In the second sampling period and when the fish were caught by hook and line much lower levels of cortisol and lactate were observed. 67

68 The results of the tagging experiments have previously been discussed in section 1 where only muscle samples were available. Here muscle cortisol levels approached five ng/ml in those fish brought on deck, those sampled in the water were relatively lower but again it is difficult to determine resting values. If one also considers figure 2.58 it is obvious that the second sampling period yielded lower plasma cortisol values than the previous sampling period in March. The last sampling appeared in July yielded the highest general plasma cortisol levels. There was a general agreement up between higher tissue cortisol levels and cortisol plasma levels (Figure 2.59) although the correlation was not particularly good. The levels in tissues for approximately one per cent of those found in plasma and probably dependent upon the amount of plasma contained within the muscle sample. Therefore tissue cortisol levels can only give a rough approximation of plasma values. Plasma Cortisol and Lactate Mazarrron 23 Plasma Cortisol Plasma Lactate 35 3 First Sampling March 23 Second Sampling May 23 Third Sampling June 23 Fourth Sampling July Z1 Z2 Z3 Z1a Z2a Z3a Z4 Z31 Z32 Z33 Z34 Z35 Z36 Z37 Z38 Z39 Z12+ Z11 Z4 Z41 Z42 Z43 Z9+ Z22+ Z27+ Z29+ Z44 Z46 Z47 Cortisol (ng/ml) and Lacate (mm) Z49 Z5 Z51 Z52 Z53 Z54 Z55 Z56 Z13+ FISH NUMBER Fig Relationship between sampling period and plasma cortisol and lactate in Mazarron Fish during sampling Cortisol Plasma v Cortisol Tissue Values Mazarron Cortisol Tissue (ng/ml) Intercept through zero y =,84x +,885 R 2 =, y =,13x R 2 =, Cortisol Plasma (ng/ml) Fig Relationships between Plasma Cortisol and Tissue Cortisol in Mazarron Fish 68

69 Catecholamines When the catecholamine levels are examined (Figure 2.6) during the four sampling phases it is evident that during the first experiments using the stretcher and the transfer net low levels of catecholamines were observed. There was a general increase in the catecholamine levels especially in phases three and four. In these phases is chasing of the fish and raising of the net was used to entrap the fish which again will lead to very high catecholamine levels. Plasma Catecholamine Levels Mazarron 23. Norepinephrine Epinephrine 35 3 First Sampling March 23 Second Sampling May 23 Third Sampling June 23 Fourth Sampling July Z1 Z2 Z3 Z1a Z2a Z3a Z4 Z31 Z32 Z33 Z34 Z35 Z36 Z37 Z38 Z39 Z12+ Z11 Z4 Z41 Z42 Z43 Z9+ Z22+ Z27+ Catechol. Conc (nm) Z29+ Z44 Z46 Z47 Z49 Z5 Z51 Z52 Z53 Z54 Z55 Z56 Z13+ FISH NUMBER Fig. 2.6 Relationship between sampling period and plasma catecholamine levels in Mazarron fish Mazarron 24 Plasma Cortisol and Lactate Mazarron Implanted Sampled Implanted Sampled Plasma Lactate (µm/ml) Cage 3 Cage Plasma Cortisol (ng/ml) ZZ11 ZZ12 ZZ13 ZZ14 ZZ15 ZZ16 ZZ17 ZZ18 ZZ19 ZZ11 ZZ111 ZZ112 ZZ113 ZZ114 ZZ115 ZZ116 ZZ117 ZZ118 ZZ119 ZZ12 ZZ121 ZZ122 ZZ123 ZZ124 ZZ125 Plasma Lactate Plasma Cortisol Fish Nr. Fig Relationships between sampling period and plasma lactate and cortisol in Mazarron fish. During the first sampling period in 24 with fish which had been tagged previously with hormonal implants the general levels of lactate and cortisol were relatively lower than in the second sampling period (figure 69

70 2.61) Considering that these fish had been disturbed by implantation of the hormones only a few days previously ( ) and cortisol levels, which in resting fish are normally below 5 ng per milliliter, were not unduly raised, shows that this type of handling was not particularly stressful for the fish. However on the second day cortisol levels in three fish ZZ 112, ZZ 113 and ZZ 124 were all raised to extremely high values. Lactate values were also higher by a factor of three fold in fish sampled during this period. The results of the Catecholamine measurements made in 24 are shown below (Fig. 2.62). Two sampling days were used and the fish nr. Indicates the chronological order of sampling and values were very variable dependent on the time taken to kill the fish. This was done by shooting the fish through the head with a large slug or a smaller calibre to shatter the backbone of the fish Plasma Catecholamine Levels Mazarron Norepinephrine Epinephrine Catechols (nm) ZZ11 ZZ12 ZZ13 ZZ14 ZZ15 ZZ16 ZZ17 ZZ18 ZZ19 ZZ11 ZZ111 ZZ112 ZZ113 ZZ114 ZZ115 ZZ116 ZZ117 ZZ118 ZZ119 ZZ12 ZZ121 ZZ122 ZZ123 ZZ124 ZZ125 Fish Nr Fig Relationship between sampling period and plasma catecholamine in Mazarron fish. If the relationship is considered between all three stress parameters, epinephrine, cortisol and lactate are considered together as shown in figure Then it is clear that a general relationship between all three parameters is present in the Mazarron fish. However, this relationship is not linear and although the general trend of increasing lactate is followed by increasing cortisol and catecholamines the timeframes of release is different for each marker. 2 Epinephrine (nm) Cortisol (ng/ml) Lactate (mm) 3 35 Fig The relationship between all three stress parameters are shown in a three-dimensional plot above. 7

71 Mazarron 25 Cortisol and Lactate Plasma Cortisol and Lactate 25 Mazarron Plasma Lactate (µm/ml) Implant Sampled Implant Sampled Cage 2 Cage 2 Cage 1 Implanted Sampled Plasma Cortisol (ng/ml) Fish Nr plasma Lactate Plasma Cortisol Fig The relationship between plasma lactate and cortisol on the three sampling days in Mazarron in 25. In cage 2 fish had been implanted two to three days prior to sampling. In cage 1 period of eight days elapsed between implantation and sampling. From figure 2.64 it appears to be obvious that the length of time between implantation and sampling had no effect on the absolute levels of cortisol and lactate although there was a tendency in cage one towards higher levels. However peak lactate levels in the plasma were lower in 25 compared with 24. The three basic criterion for stress used in this study were the levels of Cortisol, the Catecholamines Norepinephrine and Epinephrine, and Lactate in the plasma. The use of these three criteria was chosen to represent chronic and acute stress parameters. Basically cortisol levels may arise within a one to four hour time frame and then remain relatively high, Catecholamine rise relatively rapidly and also lactate levels within 3 minutes. In the following Fig. 2.65, all three variables are plotted on a three-dimensional plot as with the 24 data. Again a similar relationship as in 24 could be established without a direct linear correlation between cortisol and plasma lactate. A better correlation can be seen between catecholamines and plasma lactate. In general it was found that the levels of all three parameters were relatively lower during the sampling period in 25 compare the 24. This may be due to the better handling techniques used in the second period to prevent fish becoming too excited and avoidance of undue stress for the fish sampled. Again the technique of killing the fish in this case was much improved compared with 24 and if carried out correctly resulted in the relative instantaneous death of the fish. 71

72 16 14 Epinephrine (nm) Cortisol (ng/ml) Lactate (mm) 18 Fig The possible relationship between plasma epinephrine, cortisol and lactate are depicted in a three-dimensional plot. In a further attempt to examine the kinetics between tissue and plasma the relationship between tissue and plasma levels of lactate have been investigated. The results are shown in figure 2.66 Relationship between Tissue and Plasma Lactate Concentrations y = -,827x + 133,62 R 2 =, Tissue Lactate (mm) y = -,444x + 118,78 R 2 =, Plasma Lactate (mm) Linear (25) Linear (24) Fig 2.66 This figure compares the levels of lactate measured in plasma and tissue samples taken from the Mazarron fish in 24 and 25. It is clear that there does not appear to be a linear relationship between tissue and plasma values. In 24 and 25 tissue lactate levels were between 8 and 14 mm. This compares with plasma levels between 1 and 34 mm, indicating that there can be a difference from four to 1 fold in the concentrations between the two compartments. As a stress indicator it would appear that plasma levels are more reliable due to the fact that the tissue levels appear to be far higher almost immediately after exercise or stress. The time course of the release of lactate into the plasma and the general use of lactate within the muscle for recycling needs to be further investigated. 72

73 2.6.4 Conclusions It would appear that the three factors used for the tracking of stress in the fish were all suitable for observing differences with different handling techniques. The general premise that there are differences in the kinetics of the three factors appear to be borne out by the data shown from this study. The use of tissue lactate levels and the correlation between the three different factors certainly requires a more detailed analysis. This will be carried out during the preparation for publication of this part the final report. This is the first time to our knowledge that three-dimensional plot have been made of all three parameters, and this strategy will be used in future studies. 73

74 CHAPTER 3 WILD SAMPLING 3.1. SPATIOTEMPORAL PATTERN OF BFT SPAWNING IN CENTRAL AND EASTERN MEDITERRANEAN BASED ON GONAD HISTOLOGICAL ANALYSIS Introduction The BFT is one of the most valuable fishing resources due to its high prices on the Japanese market. In recent years, the growing demand for bluefin tuna for fattening in cages in the Mediterranean region resulted in an increased pressure on the fisheries of this species and, consequently, in a biomass reduction that has caused serious concern on the status of the eastern Atlantic bluefin tuna resource (ICCAT, 23). The knowledge of the exact location of the spawning areas as well as the spawning periods is important to manage this endangered species in terms of conservation. In fact, during the reproductive season, a remarkable concentration of adult fish occurs in the spawning areas, resulting in a high vulnerability of the breeders to the fishery. Furthermore, the improvement of the knowledge about the spawning grounds and periods of the bluefin tuna can provide helpful information to be used for designing the experimental protocols for reproduction in captivity. The criteria generally used to localise the spawning areas as well as to individuate the period of reproduction, include the presence of individuals with ready-to spawn gonads or the presence of eggs and larvae in the area. For the assessment of the reproductive state, different methods have been reported: the classification of the gonads according to arbitrary macroscopic scales (Rodríguez-Roda, 1967), the use of biometric indices like the Index of Maturity (Sarà, 1964; 1973) or Gonadosomatic Index (de la Serna and Alot, 1992); the histological analysis of the gonads (Susca et al., 21; Corriero et al., 23; Zubani et al., 23; Karakulak et al., 24a, 24b) as well as the quantitative analysis of the reproductive steroid hormones and vitellogenin (glycolipophosphoprotein precursor of the yolk proteins), whose concentrations in plasma or muscle vary according to the reproductive state of the fish (Heppel and Sullivan, 2; Bridges et al. 21; Susca et al. 21). An accurate microscopic analysis of the gonads, based on valid histological criteria, can provide precise indications of spawning periods and areas. The most reliable indicators are the presence of hydrated oocytes (a sign of imminent spawning) or the presence of post-ovulatory follicles (a sign of recent spawning) (Hunter et al., 1986). These indicate that the sampling had been carried out during the spawning period within or in the vicinity of the spawning area. The aim of this chapter is to provide further information on the spatiotemporal pattern of bluefin tuna spawning in central and eastern Mediterranean Materials and methods Gonads were obtained from 572 adult bluefin tuna (fork length, L F, ranging from 99 to 294 cm) captured between March and October 23, 24 and 25 in different Mediterranean areas (Fig. 3.1 and Tables 3.1 and 3.2) by: a) commercial vessels that made use of long lines or purse seines and operated in the North Ionian Sea (Gulf of Taranto) (12 females and 7 males), Sicilian waters (South Tyrrhenian Sea and Strait of Messina) (31 females and 12 males), waters around Malta (63 females and 4 males), Levantine Sea (between Cyprus and Turkey) (16 females and 119 males) and North Aegean Sea (14 females and 6 males); b) traditional traps operating in Sardinia Channel (2 females and 22 males). Gonad samples were also taken from 66 (39 females and 27 males) specimens farmed in floating cages in the Levantine Sea. 74

75 Fig. 3.1 Geographical location of sampling areas. a, North Ionian Sea (Gulf of Taranto), b, Sardinia Channel; c, South Tyrrhenian Sea and Strait of Messina; d, Malta; e, eastern Mediterranean (North Aegean Sea and Levantine Sea). Table 3.1 Summary of bluefin tuna ovary sampling in the central and eastern Mediterranean during 23, 24 and 25. March Apr May June July Aug Sep Oct Total 1 st half 2 nd half 1 st half 2 nd half 1 st half 2 nd half North Ionian Sea Sardinia Channel Sicily Malta Levantine Sea North Aegean (Çanakkale) Levantine Sea (farm) Total Table 3.2 Summary of bluefin tuna testis sampling in the central and eastern Mediterranean during 23, 24 and 25. North Ionian Sea Sardinia Channel March Apr May June July Aug Sep Oct Total 1 st half 2 nd half 1 st half 2 nd half 1 st half 2 nd half Sicily Malta Levantine Sea North Aegean (Çanakkale) 6 6 Levantine Sea (farm) Total Gonad slices (1-cm thick), taken immediately after the fish had been caught, were fixed in Bouin s solution or neutral 1% formalin. The samples were dehydrated in increasing ethanol concentrations, clarified in Histolemon and embedded in paraffin wax. Sections (5-µm thick) were stained with Haematoxylin-Eosin, Mallory s trichrome, and Periodic acid-shiff (Pas) reaction. 75

76 Oocyte diameters were measured on histological slides using a Quantimet 5W (Leica, Cambridge, U.K.) image analyser. The most advanced oocyte stage was recorded for each specimen as well as the presence of atretic and postovulatory follicles was also recorded. Vitellogenic atretic follicles were classified at α or β stage according to Hunter et al. (1986). Briefly, the identification of α atretic follicles was made on the basis of zona radiata fragmentation, yolk granule break down and reabsorbtion, as well as nuclear disarrangement. In β atretic follicles, yolk was completely reabsorbed, germinal vesicle disappeared and oocytes were invaded by follicular and theca cells. For females, the classification of the reproductive state was carried out according to Schaefer (1998). Female fish were classified as active when the ovary contained late vitellogenic oocytes and showed only minor atresia of vitellogenic oocytes. Active females were then classified into spawning (post-vitellogenic oocytes and/or postovulatory follicles present) and non-spawning (no evidence of recent or imminent spawning present) classes. Females were classified as inactive when the ovary contained late vitellogenic oocytes, but more than 5% of them were atretic (major atresia). Also included in this class were ovaries with pre-vitellogenic or early vitellogenic oocytes. For the classification of the reproductive state of males, the type of spermatogenic cysts was recorded and the quantity of spermatozoa in the lumen of seminipherous tubules was subjectively evaluated Results Gonad structure and development Ovary The ovary parenchyma consisted of ovigerous lamellae containing numerous follicles at different stages of development (Fig. 3.2a), embedded in a mass of connective tissue. Each developing oocyte was surrounded by a single layer of follicular cells. On the basis of oocyte morphology, staining affinity and diameter, it was possible to define the following oocyte developmental stages: Fig. 3.2 Micrographs of the ovaries from bluefin tuna specimens caught in different periods and in different areas. (a) Ovary showing ovigerous lamellae containing oocytes at perinucleolar and lipid stages. Haematoxylin-eosin staining (bar = 75 µm). (b) Oocytes at perinucleolar and lipid. Haematoxylin-eosin staining (bar = 8 µm). (c) Oocyte at lipid stage showing a thin Pas + zona radiata. PAS reaction (bar = 4 µm). (d) Ovary showing oocytes at perinucleolar, lipid, early vitellogenesis and late vitellogenesis stages. A few atretic oocytes can also be observed. Haematoxylin-eosin staining (bar = 8 µm). (e) Oocyte at early vitellogenesis stage showing PAS + cortical alveoli. PAS reaction (bar = 55 µm). (f) Particular of oocytes at early and late vitellogenesis stages. Note the increase, during oocyte development, of both zona radiata thickness and acidophilic yolk globules. Mallory s trichrome staining (bar = 4 µm). Arrowhead, perinucleolar-stage oocyte; arrow, lipid stage oocyte; asterisk, zona radiata; double arrowhead, follicular cell layer; double arrow, cortical alveolus; small arrow, nucleolus; ao, atretic oocyte; ev; early vitellogenesis oocyte; lv, late vitellogenesis oocyte; n, nucleus. 76

77 Perinucleolar stage The perinucleolar-stage oocytes (diameter µm) were polyedric cells with an intense ooplasm basophily, numerous small nucleoli adjoining the nuclear envelope, and a high nucleus-cytoplasm ratio [Figs. 3.2(a), (b), (d)]. Lipid stage The lipid-stage oocytes (diameter µm) exhibited a weak ooplasm basophily and were characterised by small lipid droplets [Figs. 3.2(a), (b), (d), 3(c)] as well as by the appearance of a thin Pas+ zona radiata [Figs. 3.2(c)]. Early vitellogenesis stage The oocytes at the beginning of vitellogenesis (diameter 22-3 µm) were characterised by small spherical acidophilic granules (yolk globules) [Figs. 3.2(d), (f)], Pas+ cortical alveoli and a 3 µm thick zona radiata surrounded by cubic granulosa cells [Figs. 3.2(e)]. Late vitellogenesis stage The oocytes at an advanced stage of vitellogenesis (diameter 3-5 µm) displayed a remarkable increase in both size and number of yolk globules [Figs. 3.2(d), (f)], as well as an emergence of Pas+ granules in the peripheral ooplasm, and an increasing in the zona radiata thickness (12 µm) [Fig. 3.2(f)]. Final oocyte maturation Post-vitellogenic oocytes (diameter µm) were characterized by: migration of the nucleus towards the animal pole associated with contemporaneous progressive coalescence of lipid droplets and yolk globules (Fig. 3.3a); germinal vesicle breakdown (Fig. 3.3b) and ooplasm hydration and complete detachment of the follicular cells (Fig. 3.3c). Fig. 3.3 Micrographs of the ovaries from bluefin tuna specimens caught in different periods and in different areas. (a) Oocyte at migratory nucleus stage. Haematoxylin-eosin staining (bar = 12 µm). (b) Ovary showing oocytes at final maturation characterised by the coalescence of lipids and yolk globules and detachment of follicular cell layer. Haematoxylin-eosin staining (bar = 6 µm). (c) Ovary showing mature (hydrated) oocytes in which the coalescence of lipid and yolk had been completed, resulting in a single yolk mass and a large oil droplet. Haematoxylin-eosin staining (bar = 4 µm). (d) Ovary from a specimen with major α atresia of vitellogenic follicles. Haematoxylin-eosin staining (bar = 14 µm). Double arrowhead, follicular cell layer; ao, atretic oocyte; l, lipid droplet; lv, late vitellogenesis oocyte; n, nucleus; h, hydrated oocyte. The presence of atretic oocytes was observed in most of the ovaries in either vitellogenic or post-vitellogenic development. A majority of atretic oocytes was found in the ovaries at the end of the spawning season. Initial 77

78 atresia of vitellogenic oocytes was characterised by ooplasm and yolk degradation and zona radiata disappearance [Figs 3.3(d)]. Testis Bluefin tuna testis (Fig. 3.4a) is constituted by seminipherous tubules, interspersed in the connective stroma, radiating from the longitudinal main sperm duct toward the testicular periphery. Within seminiferous tubules, germ cells developed in groups called germinal cysts or spermatocysts, each of them consisting of isogenic germ cells enveloped by Sertoli cell processes (Fig. 3.4b). Fig. 3.4 Micrographs of bluefin tuna testes. (a) testis showing numerous seminipherous tubules radiating from the main sperm duct toward the testicular periphery. Haematoxylin-eosin staining (bar = 5 µm). (b) Seminipherous tubules showing spermatocysts containing germ cells at different maturity stage. Rare spermatozoa can be observed in the lumen. Haematoxylin-eosin staining (bar = 4 µm).arrow, spermatogonium; double arrow, spermatocyst containing spermatocytes; md, main sperm duct; sz, spermatozoa. 78

79 Spatio-temporal pattern of gonad maturation Eastern Mediterranean During mid May-mid June, bluefin tuna caught in the Levantine Sea were reproductively active as indicated from the presence of oocytes at late vitellogenesis in the ovaries and seminipherous tubules filled with spermatozoa in the testes. Most of the females analysed displayed also oocytes in final maturation, included hydrated oocytes, as well as postovulatory follicles (Fig. 3.5). Fig. 3.5 Micrographs of the ovaries from reproductively active bluefin tuna specimens. (a) Ovary from ana active non spawning individuals with oocytes at late vitellogenesis. (b) Ovary from a spawning specimen with hydrated oocytes. (c) Ovary from a spawning specimen with both oocytes at final maturation and postovulatory follicles. (b) Mature testis with seminipherous tubules filled with spermatozoa. Haematoxylin-eosin staining (bar = 3 µm). arrow, postovulatory follicle; ev, early vitellogenic oocyte; h, hydrated oocyte; lv, late vitellogenic oocyte; p, perinucleolar stage oocyte; mn, migratory nucleus stage oocyte. Fish fattened in floating cages in the Levantine Sea and slaugthered in June were reproductively inactive as they had atretic, with major α or β atresia of vitellogenic follicles, or spent ovaries, with only unyolked oocytes (Fig. 3.6). 79

80 Fig. 3.6 Micrographs of the ovaries from reproductively inactive bluefin tuna specimens. (a) ovary showing major α atresia of vitellogenic follicles. (bar = 5 µm). (b) Ovary with unyolked occytes and advanced stage of atresia of vitellogenic follicles. (bar = 1 µm). Haematoxylin-eosin staining. Arrowhead, advanced stage of atresia of vitellogenic follicle; af, α atretic follicle. All the fish slaugthered in the period between August-October after fattening in floating cages in the Levantine Sea were reproductively inactive. Also inactive were all the fish caught in March in the North Aegean Sea. Central Mediterranean Bluefin tuna captured in April in the North Ionian Sea (Gulf of Taranto) and in the Strait of Messina (Sicily) were mostly inactive since the ovaries displayed only oocytes at perinucleolar stage or lipid stage and the testes were at early spermatogenesis stage, showing germ cells at all stages of spermatogenesis and rare spermatozoa in the lumen of seminiferous tubules. Most of the fish captured between May and early July in the North Ionian Sea, in the waters around Sicily, in Sardinia Channel and in the waters around Malta, were reproductively active. Ovaries showed oocytes at late vitellogenesis as the most advanced stage and active spermatogenesis took place in testes, being the wall of seminipherous tubules lined with meiotic and spermatidic cysts and spermatozoa abundant in the lumen of seminipherous tubules, efferent ducts and main sperm duct. Fish samples in mid-late June and during early July in the South Tyrrhenian Sea and in the waters around Malta were mostly in spawning condition. The presence of oocytes at final maturation, including hydrated oocytes, and postovulatory follicles was evidenced in the ovaries; in the testes, the lumen of seminipherous tubules, efferent and main sperm ducts were filled with spermatozoa. The presence of fish with atretic ovaries was also evidenced in the waters around Malta during early July. Only reproductively inactive fish were found in the Central Mediterranean during late June, August, September and October. 8

81 3.1.4 Discussion The knowledge of bluefin tuna spawning in the Mediterranean dates back to the early 19 s, when bluefin tuna larvae were found in Sicilian waters by Sella (1924, 1929), Sanzo (1932), Scaccini (1968) and Piccinetti and Piccinetti Manfrin, (197). In the Balearic Sea, bluefin tuna larvae were first reported by Duclerc et al. (1973). Piccinetti and Piccinetti Manfrin (1973) indicated a spawning area within the Adriatic Sea on the basis of their findings of larval and juvenile bluefin tuna, while Dicenta (1975) and Rodríguez-Roda (1975) carried out larval surveys encompassing a greater spatial range in the Spanish Mediterranean; their data showed that the waters around the Balearic Islands represent an important spawning ground for several tuna species, including bluefin tuna. Recent larval campaigns suggested the presence of other Mediterranean spawning areas. Piccinetti and Piccinetti-Manfrin (1994) reported the presence of bluefin tuna larvae all over the Mediterranean except Alboran, Ligurian, the North Adriatic seas, and the Gulf of the Lions. Nishida et al. (1998) found a high larval density in the Balearic waters, the area around Malta and the South Tyrrhenian Sea between Sicily and Calabria (Italy). The larval surveys recently carried out by the Spanish Institute of Oceanography in the Balearic Islands region confirmed the Balearic region as a spawning area for bluefin tuna, frigate tuna and albacore (Thunnus alalunga) (Garcia et al., 22). In a recent survey, bluefin tuna larvae were found by Oray and Karakulak (25) in the Levantine Sea between the coasts of Cyprus and Turkey. Recent studies based on the histological analysis of the gonads, confirmed the presence of a spawning area around the Balearic Islands, where female bluefin tuna with hydrated oocytes were found (Corriero et al., 23; Zubani et al., 23). The histological analysis performed in this work, confirmed the spawning of bluefin tuna in the South Tyrrhenian Sea between Sicily and Calabria and in the waters around Malta. The analysis of gonads from bluefin tuna collected in the Levantine Sea in the ambit of the present research project indicated the presence of a spawning area between the coasts of Turkey and Cyprus, where females with hydrated oocytes and / or postovulatory follicles were found (Karakulak et al., 24a, 24b). The cited reports indicate that bluefin tuna can spawn almost anywhere in the Mediterranean and Black Seas. However, finding a few larvae cannot account for the localization of a spawning area. The diagnostic relevance of the presence of larvae must also consider that the presence of a few larvae may be attributed to their drifting form the real spawning area or a larval misidentification may have occurred. Bluefin tuna certainly spawn in the waters around the Balearic Islands and Malta, in the South Tyrrhenian Sea (between Sicily and Calabria), and in Levantine Sea (between the coasts of Turkey and Cyprus), as indicated by both histological evidence and larval findings. The inclusion of the Adriatic Sea among the spawning areas that appears from the cited larval surveys, contradicts the observation of Corriero et al. (23) who found only individuals with atretic ovaries in this area during the spawning season. Furthermore, no evidence of the presence of a high concentration of adult bluefin tuna is reported in the literature and no fishing activities targeting bluefin tuna broodstock exist in this sea. Of course the presence of other spawning grounds within the Mediterranean sea cannot be excluded but further investigations are required to confirm this. The spawning period of bluefin tuna in the Mediterranean basin has been investigated by several authors. Sanzo (1932) indicated that bluefin tuna reproduction occurs during June and the first half of July. Rodríguez-Roda (1967) believed that the spawning period was limited to the end of June or beginning of July, since arrival bluefin tuna caught by traps during the months of May and June had gonads in the prespawning stage, while return fish caught in July and August were in the post-spawning stage. Sarà (1964, 1973) suggested a spawning period during mid June early July. According to de la Serna and Alot (1992), the most intense reproductive activity in the western Mediterranean occurred in July. Moreover, data from Susca et al. (21), based on histological and immunohistochemical investigations as well as sex steroid and vitellogenin plasma level analyses, stated that bluefin spawning occurs after mid-june. Medina et al. (22), based their studies on the histological analysis of the ovaries and classified all the 24 female bluefin tuna sampled around the Balearic Islands from the 26th June to the 2nd July to be in the spawning stage. Oocytes in final maturation (post-vitellogenesis) and post-ovulatory follicles were observed by Sarasquete et al. (22) in the ovaries of bluefin tuna caught in July in the western Mediterranean. Corriero et al. (23) found spawning females in the Balearic waters in late June - early July. The data reported in the present report show that bluefin tuna even spawn in Maltese waters in the same period, whereas spawning in the Levantine Sea occurs almost one month earlier than in the central and western Mediterranean Sea (Karakulak et al., 24a, 24b). This anticipation can be ascribed to the fact that temperatures suitable for bluefin reproduction (around 24 C) are reached in the Levantine Sea earlier than in the rest of the Mediterranean area. 81

82 3.2 AGE AND GROWTH OF BFT FROM THE MEDITERRANEAN SEA Introduction The bluefin tuna (Thunnus thynnus L.), one of the most highly developed among the tuna species" (Cort and Liorzou, 1991), is one of the fastest, largest and most wide-ranging of animals. It can grow up to 7 kg, travel on trans-oceanic migrations and swim at 9 km per hour (Safina, 1993). The Atlantic bluefin tuna (Thunnus thynnus thynnus) distribution extends over an extraordinarily large area ranging off the Atlantic coats of Europe and Africa, from the North Cape to the Hope Cape, and off the American coasts from Newfoundland to 4 S latitude and in most of the intervening oceanic areas (Mather et al., 1995). The International Commission for the Conservation of Atlantic Tunas (ICCAT) regulates bluefin tuna fishery and currently recognizes two stocks, primarily on the basis of the location of their spawning sites: the west and the east Atlantic stocks (separated by 45 W meridian) (Sissenwine et al., 1998; Nemerson et al., 2). Recent evidence indicates that the two bluefin tuna populations overlap on North Atlantic foraging grounds (Block et al., 25). The main objective of age and growth studies in fish is to estimate the mean size of each age class and determine the growth parameters (Mather et al., 1995). The growth rate of fish is an essential component of models used in stock assessment of fish population and small variations in growth rates can have significant impact on the outcomes of the population analysis (Megalofonou, 2). This is particularly important in the case of the Atlantic bluefin tuna since the actual existence of two stocks is a matter of debate. There are several ways of modelling the growth of fish (Ricker, 1975). However, the von Bertalanffy growth function has been by far the most studied and most used of all length-age models in fish biology (King, 1995). To age bluefin tuna, a variety of methods based on the size analyses of caught individuals or interpretation of the discontinuities of different calcified structures (hard parts) of the fish have been used. The reading of hard parts, like otoliths, scales, spines and vertebrae is based on the number of marks, usually called annuli, which are interpreted as periodic events (Sella 1929; Mather and Schuck 196; Compeàn-Jimenez and Bard, 1978; Megalofonou and De Metrio, 2). In tuna species, among the different hard parts used for growth studies, fin spines have been often chosen since they are easy to sample and well-defined growth marks are evident on them (Megalofonou, 2; Megalofonou and De Metrio, 2). The objective of this study is to increase the exiting mass of data on length-age correlation and growth rate of Atlantic bluefin tuna from the Mediterranean Sea and to verify the possible existence of a differential growth between males and females Materials and methods Sampling The first spiniform ray of the first dorsal fin was taken from 757 bluefin tuna captured between 1998 and 25 in different locations of the central Mediterranean Sea (North Ionian, South Adriatic and South Tyrrhenian seas and waters around Malta). The fish were caught by commercial vessels which made use of long-line or purse seines. For each fish sampled, the fork length (FL) and the round body weight (RW) were measured to the nearest cm and 1 g, respectively. Sex was determined by macroscopic observation of the gonads and subsequently confirmed by histological analysis Age and growth estimation The age was determined using the technique described by Cort (1991) and Megalofonou (2). Three serial cross-sections about.7 mm thick were obtained from each spine at the point near the condyle base using a low speed saw and diamond wafering blades. The sections were mounted with Eukitt Mounting Medium (Electron Microscopy Sciences, Hatfield, PA, U.S.A.) on glass slides and observed with a binocular lens microscope Wild M3C (Leitz, Heerbrugg, Switzerland) under transmitted light connected trough a digital camera DC 3 (Leica, Wetzlar, Germany) to the image analyser Quantiment 5 W (Leica, Wetzlar, Germany). Interpretation of growth band was based on the recognition of the narrow translucent and wider opaque zones that are assumed to represent slow and fast growth, respectively. The number of translucent zones or rings interpreted as annual events was counted in order to assign an estimated age to the fish and build a size-age key. 82

83 As the nucleus of the spine is reabsorbed and the first rings begin to disappear at age 3, the mean diameter of the first rings of younger specimens was used to date the first ring of older specimens (Corriero et al., 25). Two reading of each spine were made independently by one reader. When there was disagreement between counts of translucent bands, spine was read again for a third time. To estimate previous growth history of bluefin tuna by back calculation, the relationship between the spine radius (R) and the fork length was studied. Linear and curvilinear equations were tested using regression analysis. The radius of the spine was defined as the distance between the estimated centre of the cross section and the edge of the section (Compeàn-Jimenez & Bard, 1983). Back calculation of FL at estimated age was obtained using the formula of Tesch (1971) and Ricker (1975). Estimates of theoretical growth in length were obtained by fitting the von Bertalanffy growth model to the mean lengths at estimated ages using the Sparre (1987) method. Theoretical growth in weight was obtained by converting length to weight using the length-weight relationship. Linear regression was used to determine length-weight relationship. The quantity φ =lnk+2lnl, was computed (Sparre et al.,1989) to compare bluefin tuna growth parameters of this study with those estimated in other studies from the Atlantic Results Fork length and round weight of sampled bluefin tuna ranged from 51. to 255. cm and from 2.6 to 247. kg, respectively. The fork length-round weight relationships of males and females were (Fig. 3.7 and 3.8): males: W R = 5 x 1-5 x FL 2.76 (r 2 =.99) N = 338 females: W R = 3 x 1-5 x FL 2.85 (r 2 =.98) N = 299 WR (kg) FL (cm) Fig. 3.7 Fork length-round weight of male bluefin tuna, sampled in the Mediterranean Sea during N =

84 WR (kg) FL (cm) Fig. 3.8 Fork length-round weight of female bluefin tuna, sampled in the Mediterranean Sea during N = 299 Both regressions were highly significant (P<.1). The slopes of the regressions indicated an allometric growth. The two sexes of bluefin tuna had significantly different slopes (t=2.778; P<.1) indicating heterogeneity of the coefficient with sex. The spine radius-fork length relationship did not differ between sexes (t=1.687; P<.1) and then a pooled significant linear regression (Fig.3.9) was developed with the equation: FL = x R (r 2 =.97) N = FL (cm) R (mm) Fig Relationship between fork length (FL) and spine radius (R) of the first dorsal fin for bluefin tuna sampled in the Mediterranean Sea between 1198 and 25. N = 344. A summary of the measurements obtained from spine reading (ring diameter) at age is given in table

85 Table 3.3. Summary of the measurements obtained from spine reading (ring diameter) at age for bluefin tuna captured in the Mediterranean Sea between Age Mean Min Max St. dev. N I II III IV V VI VII VIII IX X XI XII XIII XIV XV Base on the counts of the translucent zones (Fig. 3.1), the estimated age ranged between 1 to 15 years. Especially, the estimated age for males ranged between 1 to 15 years while the range for females was from 1 to 14 years. Mean fork lengths at estimated ages were determined by sex and the von Bertalanffy growth model was fitted to these with 95 % CI (Tables 3.4 and 3.5; Fig. 3.11). The comparison among the observed lengths at age, those back calculated and those predicted by von Bertalanffy model are reported in Tab Fig. 3.1 Images of bluefin tuna spine sections. (a) Age 1 specimen with 58 cm FL. One ring is visible. (b) Age 5 bluefin tuna with 138 cm FL. Four rings are visible and one was reabsorbed. Arrows indicate visible rings. Magnification bars = 2 mm. 85

86 Tab. 3.4 Mean observed fork length at age and those predicted by the von Bertalanffy model, for bluefin tuna Thunnus thynnus sampled in the Mediterranean Sea. Round weights at age derived from the conversion of the predicted fork lengths using the length-weight relationship. Mean observed FL (cm) Predicted by von Bertalanffy model Estimated FL (cm) RW (kg age group Males Females Males Females Males Females Tab. 3.5 Summary of parameter estimates for the von Bertalanffy growth equation on fork length (cm) of bluefin tuna Thunnus thynnus sampled in the Mediterranean Sea. Sex Parameter Estimate Lower 95% CI Upper Sexes combined Males Females L k t L k t L k t FL (cm) males females total Age (year) Fig Von Bertalanffy growth curves for sexed and pooled bluefin tuna Thunnus thynnus from the Mediterranean Sea. 86

87 Tab. 3.6 Comparison among mean observed, back-calculated and predicted by the von Bertalanffy model, fork lengths (FL) at age for bluefin tuna sampled in the Mediterranean Sea (sex combined). Estimated age group Mean observed FL (cm) FL predicted by von Bertalanffy model (cm) Back- calculated FL (cm) The growth parameters of bluefin tuna from the Mediterranean and Atlantic as well as the calculated quantityφ seem to show slight differences (Table 3.7). Table 3.7 Comparison of growth parameters of Atlantic and Mediterranean bluefin tuna Thunnus thynnus, estimated by different authors using different methods (φ =lnk+2lnl ) Author Sex Ageing Growth parameters methods L k t Area φ Longevity (years) Rodriguez-Roda (1971) combined vertebrae from Mather and Sakagawa combined vertebrae Schuck and Coan from Mather and (1974) combined vertebrae Jones Butler et al. (1977) males otoliths females otoliths Bard et al. (1978) combined vertebrae Parrak and Phares (1979) Compéan-Jimenez and Bard (198) Farrugio (198) Farber and Lee (1981) Hurley et al. (1981) Compéan-Jimenez and Bard (1983) combined markrecapture data combined spines combined combined length frequency markrecapture data Northeast Atlantic Northwest Atlantic Northwest Atlantic Northwest Atlantic Northwest Atlantic Northwest Atlantic Northwest Atlantic Northeast Atlantic Mediterranean combined vertebrae males females length frequency length frequency combined spines Northwest Atlantic Northwest Atlantic Northwest Atlantic Northwest Atlantic Northeast Atlantic

88 Hurley and Iles (1983) Northwest males otoliths Atlantic females otoliths Northwest Atlantic Nagai (1985) combined length Northwest frequency Atlantic Cort (1991) combined spines Northeast Atlantic Present study combined spines Mediterranean males spines Mediterranean females spines Mediterranean Discussion In the present study we reported the length-weight relationship, the spine radius-fork length relationship, the mean fork lengths at estimated age, and the von Bertalanffy equation for bluefin tuna captured in the Mediterranean Sea. The length-weight relationship, calculated separately for males and females, confirm a higher weight at length for males previously reported for bluefin tuna captured in the Mediterranean (Tawil et al., 24). The spine radius-fork length relationship did not differ between sexes. The high correlation coefficient indicates a strong relationship between the two variables, which justifies the use of the spine radius size at age for making back calculation of previous growth (González-Garcés and Fariña-Perez, 1983). Observed lengths at age, those back calculated and those predicted by von Bertalanffy, reported in the present paper, are almost overlapping. The back calculation was already reported for eastern Atlantic bluefin tuna by Rodríguez-Roda (1964), on the basis of vertebral rings. The back calculated lengths reported by Rodríguez- Roda (1964) were considerably less than those obtained directly from fish of the same ages, probably due to the time interval between ring deposition (winter) and sample collection (spring-summer). The observed mean fork lengths at age, reported in the present study, indicate that males grow slightly faster than females and reach a slightly larger size. This finding has already been reported for western Atlantic (Hurly and Iles, 1983), but never for eastern Atlantic bluefin tuna. The observed mean fork lengths at age were very close to those predicted with the von Bertalanffy equation, thus indicating the reliability of the spine readings performed in the present work. The von Bertalanffy parameters reported in this study, indicate a theoretical maximum length of 382 and 349 cm FL, for males and females respectively. Therefore, the larger body size reached by males is due both to a faster growth and a major longevity. The von Bertalanffy parameters and the consequent φ here reported are in reasonable agreement with those calculated with different methods both for eastern and western Atlantic bluefin tuna (Tab. 3.7). In conclusion, the present study confirms the data already available in the literature regarding eastern Atlantic bluefin tuna age and growth, and indicates that male specimens attain a larger size, grow faster and are have a major longevity than females SIZE AND AGE AT SEXUAL MATURITY OF FEMALE BFT FROM THE MEDITERRANEAN SEA Introduction The Atlantic bluefin tuna (Thunnus thynnus Linnaeus, 1758) is an important fishing resource in the Atlantic Ocean and the Mediterranean Sea. This species has been considered to be overexploited since the 198 s (Sissenwine, 1998). The International Commission for the Conservation of Atlantic Tunas (ICCAT) regulates this fishery and currently recognizes two stocks, the west and the east Atlantic stock (separated by 45 W meridian), the latter including the Mediterranean Sea. The western Atlantic population spawns in the Gulf of Mexico and in the Florida straits in April - July (Richards, 1976; Montolio and Juarez, 1977; Rivas, 1978; Baglin, 1982); whereas the eastern Atlantic population spawns in the Mediterranean during May - July (Rodríguez-Roda, 1967; Susca et al., 21; Medina et al. 22; Corriero et al., 23; Karakulak et al., 24; Oray and Karakulak, 25). Western Atlantic bluefin tuna mature at the age of 6 and are considered fully 88

89 mature by the age of 8, at a weight of 135 kg (Baglin, 1982; NRC, 1994). On the other hand, according to Rodríguez-Roda (1967), eastern Atlantic bluefin tuna mature at the age of 3, at a weight of 15 kg and are fully mature by the age of 5. Knowledge of first sexual maturity has important implications for stock management and regulation of the fishery. The aim of this paper is to accurately indicate the size and age of first sexual maturity for female eastern Atlantic bluefin tuna Materials and Methods Sampling Ovary and spine samples were collected from 51 bluefin tuna from May to September between 1998 and 24 in the waters around Balearic and Malta Islands, in the South Adriatic, North Ionian, South Tyrrhenian and Northern Levantine seas and in the Sardinia Channel. The fish were caught by commercial vessels which made use of long-lines, drift nets and purse seines and also traditional traps (tonnare) operating in Sardinia, Italy. For each fish caught fork length (FL) was measured to the nearest cm and the date and place of capture were recorded. Ovary samples were fixed in Bouin s solution or 1% buffered formalin prior to histological analysis. For age determination, the first spiniform ray of the first dorsal fin was taken and stored at 2 C Ovary histology and reproductive state Ovary samples were dehydrated in increasing ethanol concentrations, clarified in Histolemon and embedded in paraffin wax. Sections were cut (5 µm thickness) and stained with haematoxylin eosin. The oocyte developmental stages were classified according to Corriero et al. (23) and the reproductive state was assessed following Schaefer (1998). Oocyte atretic stages were classified according to Hunter and Macewicz (1985). On the basis of the classification scheme used, the distinction between immature and mature inactive fish was based on the presence of atresia of vitellogenic follicles, a sign of past reproductive activity. As previously reported (Corriero et al., 23), no sign of atresia were observed in bluefin tuna captured some months after the reproductive season, due to their complete re-absorption. For the present study, samples collected during a temporal window from May to September only were used. This periodic sampling allowed a clear distinction between mature and immature specimens Size at first sexual maturity The body length at median sexual maturity (L 5 ) was estimated by fitting a logistic function to the fraction of mature fish per 5 cm FL intervals by nonlinear regression using the FISHPARM program (Saila et al., 1988). L 5 was defined as the smallest length interval in which 5% of the specimens were mature Age determination The age was determined for all the fish belonging to the size classes corresponding to L 5 (n = 2) and L 1 (the size for which all the fish were mature; n = 4) using the technique described by Cort (1991) and Megalofonou (2). Briefly, three serial cross-sections about.7 mm thick were obtained from each spine at the point near the condyle base using a low speed saw and diamond wafering blades. Spine sections were observed with a binocular lens microscope under transmitted light connected to the image analyser Quantimet 5W (Leica, Cambridge, U.K.). Interpretation of growth bands was based on the recognition of the narrow translucent and wider opaque zones that are assumed to represent slow and fast growth, respectively. Therefore, the number of translucent zones or rings, interpreted as annual events, was counted in order to assign an estimated age to the fish. As the nucleus of the spine is reabsorbed and the first rings begin to disappear from age 3, the mean diameter of the first rings of younger specimens was used to date the first visible ring of older specimens (Rodríguez-Marín, 24). Two readings of each spine were made independently by one reader. When there was disagreement between counts of translucent bands, spines were read again for a third time Results Ovary histology and reproductive state On the basis of the ovary histological pattern (Fig ), 57 individuals or 11.3% of the specimens analysed were immature whereas 444 or 88.7% were mature. 89

90 Fig Micrographs of the ovaries from bluefin tuna specimens captured in the Mediterranean Sea. Haematoxylin-eosin staining. (a) Ovary from an immature fish showing only perinucleolar stage oocyets (bar = 1 µm). (b) Ovary from an active non-spawning specimen with late vitellogenic oocytes (bar = 5 µm). (c) Ovary from a spawning bluefin tuna with hydrated oocytes (bar = 5 µm). (d) Ovary from a spawning individual with both hydrated oocytes and post-ovulatory follicles (bar = 5 µm). Inset: higher magnification of a post-ovulatory follicles (bar = 3 µm). (e) Ovary from an inactive mature specimen characterized by major α atresia of vitellogenic follicles. (bar = 5 µm). (f) Ovary from an inactive mature fish showing perinucleolar stage oocytes and δ atretic follicles (bar = 1 µm). Arrow, post-ovulatory follicle; arrowhead, δ atretic oocyte; af, α atretic follicles; h, hydrated oocyte; lv, late vitellogenic oocyte Body lengths at sexual maturity Frequencies of immature and mature ovaries in different FL groups are shown in Fig No mature fish were found below 1 cm FL. The estimated body length at median sexual maturity (L 5 ) was 13.6 cm FL (S.E =.99). Fittings of the logistic model resulted in: Y = 1/(1+EXP(-.1739(X-13.6)) Where Y = % mature; and X = body size (FL) All the fish above 135 cm FL were found to be mature. 9

91 1 9 Percent sexually mature L 5 = 13.6 n = 51 r 2 = Fork length (cm) Fig Percent mature female bluefin tuna by 5-cm fork length interval, fitted to a logistic function. Arrow indicates body length at median sexual maturity (L 5). n = sample size Age at sexual maturity Twenty specimens constituted the size class that contained L 5 (1-14 cm FL), while 4 fish were included in the size class beyond which 1% maturity was reached ( cm FL). All the spines analysed showed the complete formation of the ring corresponding to their last year of life. Among the 2 fish included in the 1-14 cm size class, 16 belonged to the age group 3 (Fig. 3.14a) and 4 to the age group 4 (Fig. 3.14b). The 4 fish contained in the cm size class belonged to the age group 4 (4 specimens) and 5 (36 specimens) (Fig. 3.14c). 91

92 Fig Images of bluefin tuna spine sections. (a) Age 3 specimen with 13 cm FL captured on the 7 July; two rings are visible and one ring was reabsorbed. (b) Age 4 fish with 14 cm FL caught on the 16 May; three rings are visible and one was reabsorbed. (c) Age 5 bluefin tuna with 138 cm FL sampled on the 19 May; 4 rings are visible and one was reabsorbed. Arrows indicate visible rings. Magnification bars = 2 mm Discussion The present paper represents the first attempt to determine the size at sexual maturity for the eastern Atlantic female bluefin tuna using a method based on the statistic elaboration of data coming from histological analysis (Lowerre-Barbieri et al., 1996; DeMartini et al., 2). On the basis of the macroscopic evaluation of the ovary maturity stage, Rodríguez-Roda (1967) estimated that 5% of the female bluefin tuna of the eastern stock are reproductively active at a size of 97.5 cm FL, while 1% maturity is reached between 115 and 12 cm FL. Tawil et al. (21), in a preliminary approach 92

93 to the study of sexual maturity based on the histological analysis of the ovaries of 21 bluefin tuna, found mature specimens above 115 cm FL. Further approximate information on the first sexual maturity of the eastern Atlantic bluefin tuna comes from investigations carried out for different aims. In a stereological study on bluefin tuna fecundity, Medina et al. (22) reported that the smallest mature female sampled in the Balearic waters was 116 cm FL. During a histological description of the ovarian cycle, mature females over 11 cm FL were found (Corriero et al. 23). The analysis of the spines of fish captured between May and September indicate that all the specimens had completed the formation of the ring corresponding to their last year of life. This finding is in agreement with Cort (1991) and Megalofonou and De Metrio (2) who reported that ring completion occurs during April and May for bluefin tuna caught in the Mediterranean. Our data indicates that the estimated age of most of the specimens with FL = L 5 is 3 years while 1% maturity is reached at 4-5 years. The age estimates reported in the present work are consistent with previous studies regarding age and growth of eastern Atlantic bluefin tuna carried out by the count of translucent zones in the dorsal spines. Cort (1991) and Megalofonou and De Metrio (2) found bluefin tuna with 1 FL < 15 cm belonged to the age group 3; Cort (1991) reported that most of the analysed fish with 13 FL < 135 cm were 5 years old. Although to our knowledge there is no study reporting the size at 5% sexual maturity for the Western Atlantic bluefin tuna, the available data indicate that in this population, maturation starts at age 6 and 1% maturity is reached by age 8 at a size of 19 cm FL (Baglin, 1982; NRC, 1994). The evidence reported in this paper confirms that the eastern Atlantic female bluefin tuna have a size and age of first sexual maturity that is markedly lower than the western Atlantic stock. This represents further evidence that leads to the scientific correctness of the separated management of the two stocks REPRODUCTIVE PERFORMANCE OF BFt IN THE EASTERN ATLANTIC AND WESTERN MEDITERRANEAN: INFLUENCE OF SAMPLING GEARS IN THE ASSESSMENT OF REPRODUCTIVE PARAMETERS Introduction Modern electronic tagging programs have provided invaluable data on the dynamics of Atlantic bluefin tuna populations (Lutcavage et al. 1999, 2, Block et al. 21, 25, De Metrio et al. 22, 23, Wilson et al. 25), but this information need to be accompanied with meaningful biological interpretations of migratory patterns. Accurate assessment of reproductive parameters, such as age at maturity, fecundity, spawning frequency, and spawning sites and schedules, is essential for the understanding of the population dynamics of tunas and the management of tuna fisheries. Considerable progress has been made in recent years in the understanding of the reproduction of Atlantic bluefin tuna, but many questions still remain to be answered. One of the most controversial issues regarding bluefin tuna reproductive biology is the age at maturity. While age at 5% maturity has been definitely established at 3 years for eastern populations (Corriero et al. 25), maturity is believed to start as late as age 6 in the western stock (Baglin 1982). Such a difference in the age of maturity between eastern and western populations is difficult to support in terms of reproductive physiology and ecology, hence further investigations on bluefin tuna reproduction in the West Atlantic and Mediterranean are needed with a view to clarify these and other discrepancies. The list of classical Mediterranean spawning sites (Balearic archipelago and southern Thyrrhenian Sea) has been recently extended with the addition of two spawning grounds located in the Levantin Sea (Karakulak et al. 24, Oray & Karakulak 25) and around Malta (Corriero, pers. comm.). Understanding and quantifying the reproductive potential of bluefin tuna local populations in all these geographic locations is of great importance to research into population dynamics and management models for the species. A major limitation of biological research on wild bluefin tuna populations is the nearly utter dependence of scientific sampling on commercial fisheries. The inability to design the most appropriate sampling protocols and methods of capture may result in biased interpretations associated with the particular selectivity of the fishing gear used. In the present study, a comparative histological and stereological analysis of gonads from bluefin tuna caught by longline and purse seine vessels was undertaken with the aim to (1) improve our knowledge on the reproductive biology of bluefin tuna in the western Mediterranean breeding area, and (2) evaluate the influence of the methods of capture in the assessment of reproductive parameters Materials and Methods Sample collection 93

94 Eighty five migrant bluefin tuna were caught by trap off Barbate and Zahara de los Atunes (Cádiz, southern Spain) in 23, 24 and 25 as they entered the Mediterranean Sea to spawn. The sample of each of these years consisted of 2 sub-samples, collected at late April-early May and late May-early June (beginning and end of the trap fishing season, respectively). Bluefin tuna breeders were sampled from waters around the Balearic Islands by longline and purse-seine vessels. A total number of 223 individuals, 14 females and 83 males, were collected by Japanese-type longline between May and July in 23, 24 and 25. Their estimated weight averaged 21.5 ± 52.1 kg (range: kg). The longlines bore up to 1,2 hooks baited with squid (Illex sp), which were distributed in the water column at depths ranging between 5 and 125 m approximately, with water temperatures between 15-17ºC and 14ºC, respectively. The lines were always retrieved at dawn, usually hours after they were set. The purse-seine sample was obtained between mid June and mid July in 2, 21, 22 and 25, and comprised a total number of 99 tuna, 56 females and 43 males, weighing ± 83.7 kg (range: kg). This sample was far smaller than the longline sample because purse seine-caught tuna are maintained alive for farming, thereby only the few specimens that die during fishing operations or transfer to towing cages are available. All individual tuna were measured to the lowest 1 cm curved fork length (CFL); this measure was then transformed into straight fork length (FL) using the formula FL = CFL.955 for further estimations of total body weights (BW), according to the equation BW = FL 3.92 (ICCAT conversion factors: Following evisceration, the gonad pairs were weighed to the nearest 1 g, and the gonadosomatic index (GSI) calculated as GSI = GW 1 / BW, where GW represents the gonad weight. Volumetry of gonads is frequently difficult to be carried out aboard commercial fishing vessels. So, when ovarian volume (OV) measures were unavailable OV was estimated from the ovarian weight (OW) using the equation OV =.9174 OW, which resulted from a previous regression analysis between OW and OV values measured in a sample of 141 individuals whose OW ranged between.54 and 7.29 kg (r 2 =.98) Histology A transverse slice from the middle region of one of the gonads (approximately 5-mm thick) was cut into smaller pieces, each comprising several lamellae, which were fixed and preserved in 4% neutral buffered formaldehyde until use. Once in the laboratory, the tissue samples were dehydrated in a graded series of ethanol, cleared in xylene and embedded in paraffin wax. The volume loss occurred throughout dehydration of the ovarian tissue in 28 specimens was estimated to be close to 25%, with V F =.72 V I (r 2 =.89; n = 28), where V F is the volume of the histological specimen measured after dehydration and V I represents its initial volume. Cross histological sections were cut at 6 µm, and testicular samples stained with haematoxylin-eosin (H-E), whereas ovarian sections were stained with haematoxylin-vof (H-VOF) (Gutiérrez 1967). H-VOF staining was preferred over conventional H-E staining for ovarian tissue because it aids identifying tiny yolk granules in early vitellogenic oocytes, thus allowing neat discrimination between this population of oocytes and large previtellogenic oocytes (lipid stage), which have similar size and morphological appearance. Slides were examined and photographed on a Leitz DMR BE light microscope equipped with a digital camera Male histological classification Male tuna are considered mature (Fig. 3.15) when spermatozoa are present in the sperm duct. Schaefer (1996, 1998) proposed a further classification for mature males based on histological evidence of recent spawning, which is detectable only within the 12 hours following the spawning event. Unfortunately, accurate determination of the spawning condition of males was impracticable in this study because the exact time of captures was unavailable, and some histological specimens did not include the main sperm duct Oocyte and female classification Bluefin tunas show an asynchronous pattern of ovarian development (Baglin 1982, Schaefer 21, Susca et al. 21, Corriero et al. 23, Abascal & Medina 25). Hence, all stages of oocyte development can be found in the ovary during the reproductive season. According to the different developmental stages, oocytes were classified as perinucleolar, lipid stage, vitellogenic, migratory nucleus, and hydrated (Fig. 3.15). Additionally, histological sections of the ovaries were examined for the presence of atretic and postovulatory follicles. Although vitellogenic oocytes are commonly divided into early yolked and advanced yolked (Schaefer 21), we have grouped all these oocytes into a single class because no clear borderline can be objectively traced between two distinct categories of vitellogenic oocytes, which may cause problems in stereological quantifications. Depending on the most advanced group of oocytes encountered in the ovary and the extent of atresia, and following classifications by Schaefer (1996, 1998) and Schaefer et al. (25), 94

95 bluefin tuna females were classed into one of four maturational stages (Fig. 3.15). The ovaries of active nonspawning females contain advanced vitellogenic oocytes and minor, if any, α atresia (Fig. 3.15C). Active females are classified as active spawning if the ovaries show additional evidence of either recent spawning (postovulatory follicles are present) or imminent spawning (migratory-nucleus or hydrated oocytes can be identified) (Fig. 3.15D). Females are considered inactive mature when they have entered into regression following a phase of reproductive activity, in which case the ovary contains either previtellogenic or early yolked oocytes plus α and/or β atresia (Fig. 3.15E), or advanced yolked oocytes plus major atresia. Finally, in immature females only previtellogenic or early yolked oocytes and no sign of atresia are found in the ovary (Fig. 3.15F) Image analysis and stereology Ten different images of each ovarian histological section were recorded. A calibration scale was photographed at the same magnification in order to accurately determine cell dimensions in the digital images. The number of oocytes of different categories present in the ovaries was estimated by the stereological method (Weibel & Gómez 1962, Weibel et al. 1966), which has been applied for oocyte counting in various fish species (Emerson et al. 199, Greer Walker et al. 1994, Coward & Bromage 1998, 21, 22a,b, Bromley et al. 2, Medina et al. 22, Murua et al. 23, Cooper et al. 25). The stereological procedure allows to calculate the number of oocytes per unit volume of ovary (N V, numerical density) from microscope images, according to the formula: N V K N = β V 3/ 2 A 1/ 2 V, where β is a shape coefficient, K is a size distribution coefficient, N A is the number of oocyte transections per unit area, and V V is the partial area of oocytes of a given category (volume fraction or volume density). The total number of oocytes contained in the gonads is then readily obtained extrapolating N V to the whole ovarian volume. We did not count perinucleolar oocytes, since the size of this stock has been found to remain invariable at about 6,5 cells per gram of BW throughout the reproductive season (Medina et al. 22). The procedure employed for stereological analysis was as described by Medina et al. (22) with a slight modification. In the previous study V V was calculated according to the original methodology, that is, by counting the number of points of a Weibel grid that overlaid the transections of the considered oocyte type on printed micrographs. In the present investigation we have determined V V by image analysis of digital micrographs using the software ImageJ ( Both procedures produced equivalent results Statistical analysis Because individual sizes were found to be correlated with fecundity, to test differences between both groups, data of GSI, V V, N V, absolute number of oocytes, and relative number of oocytes were compared by analysis of covariance (ANCOVA), with BW as the covariate. When relevant, Dunn-Sidak multiple comparison tests were performed using SPSS statistical software. A p-value <.5 was considered statistically significant for all tests Results Gonadosomatic indices Figure 3.15 shows the condition indices recorded in bluefin tuna caught in the Strait of Gibraltar area at the beginning and at the end of the eastward migration (23, 24 and 25). A significant increase in GSI occurred in all years as the migration period advanced. Significant differences were also found in the fat body weight, whose volume decreased as the gonad matured. The liver mass increased slightly in females, and showed an irregular trend in males. 95

96 GSI April GSI June FBI April HSI June HSI April GSI Early May GSI Late May HSI Late May HSI Early May GSI late May GSI early May HSI late May HSI early May.4.2 FBI June Barbate 23 - Indices FEMALES FBI Early May FBI Late May Barbate 24 - Indices FEMALES.4.2 FBI early MayFBI late May Barbate 25 - Indices FEMALES GSI June GSI Late May GSI April HSI April FBI April FBI June Barbate 23 - Indices MALES HSI June GSI Early May HSI Early May HSI Late May FBI Early MayFBI Late May Barbate 24 - Indices MALES GSI early May Barbate 25 - Indices MALES HSI late May Figure 3.15 Organ indices in tuna from the Strait of Gibraltar recorded at the beginning and end of the migratory period. Years 23 (A), 24 (B) and 25 (C). Longline-caught male tuna showed significantly smaller gonads (GSI = 2.87 ± 1.49) than specimens caught by purse-seine (GSI = 4.27 ± 1.71) (Table 3.9). A different gonad development was also found between females caught by trap (GSI = 1.1 ±.4) and longline (GSI = 2.83 ± 1.39) (Table 3.8), as well as between longline and purse-seine (GSI = 3.85 ± 1.66). We found no interannual significant differences in GSI values within groups in any of the different sampling gears. Table 3.8 Mean GSI and stereological data from bluefin tuna fished in Baleares by long-line in are compared with those sampled at Barbate trap (ANCOVA, P< 5). Barbate (23-25) Baleares Longline (23-25) GSI 1.1 ±.4 * 2.8 ± 1.4 Atretic oocytes (V V ).5 ±.5.7 ±.7 Lipid stage oocytes No. per g of BW (g -1 ) 54 ± 45.2 * ± 62.4 Vitellogenic oocytes No. per g of BW (g -1 ) 73.5 ± 59.7 * ± Maturation stage oocytes No. per g of BW (g -1 ) ± * 2.1 ±

97 Table 3.9 GSI and stereological data (mean ± SD) from bluefin tuna caught by longline and purse seine (n: number of individuals examined). Differences between the two groups of means were significant (ANCOVA, p <.5) where indicated (*) Longline Purse Seine Males GSI Females GSI 2.87 ± 1.49 (n = 82) 2,83 ± 1,39 (n = 138) 4.27 ± 1.71* (n = 42) 3.85 ± 1.66 * (n = 56) Lipid-stage oocytes V V,6 ±,2 (n = 119) Nv (ml -1 ) ( 1 3 ) ± 17.6 (n = 119) No per individual ± ( 1 6 ) (n = 116) No. per g of BW ± (g -1 ) (n = 116) Vitellogenic oocytes V V,26 ±,12 (n = 119) Nv (ml -1 ) ( 1 3 ) 9.7 ± 4.7 (n = 119) No per individual ± ( 1 6 ) (n = 116) No. per g of BW ± (g -1 ) (n = 116) Final-maturation oocytes V V.1 ±,3 (n = 119) Nv (ml -1 ) ( 1 3 ).6 ±,23 (n = 119) No per individual.3 ± 1.22 ( 1 6 ) (n = 116) No. per g of BW 1.65 ± 6.5 (g -1 ) (n = 116) Atretic oocytes V V,7 ±,7 (n = 119).7 ±.4 * (n = 49) ± (n = 49) ± (n = 48) ± * (n = 48).24 ±.11 (n = 49) 7.71 ± 3.68 * (n = 49) 34.4 ± 36.4 (n = 48) ± (n = 48).19 ±.14 * (n = 49) 2.6 ± 2.23 * (n = 49) 9.47 ± * (n = 48) ± * (n = 48).4 ±.6 * (n = 49) In order to analyse possible temporal differences in longline-caught tuna, the whole sample was split into two groups corresponding to early captures carried out in May through mid June, and later catches occurred between mid June and mid July. This latter group encompasses longline catches that co-exist temporally with the purse-seine fishery. Multiple comparison test following the ANCOVA analysis showed that GSI remained unchanged throughout the whole fishing campaign in females caught by longline (Table 3.1). This appears to indicate that dissimilarities in the degree of ovarian development observed between longline and purseseine samples are largely due to markedly different selectivity of the gears rather than temporal differences. For males, however, strongly different GSI values were found between the specimens caught by longline in May-mid June and those caught by purse seine, whereas the tuna fished by longline between mid June and 97

98 mid July exhibited intermediate GSI values that were not statistically different form either of the two other groups (Table 3.1). Table 3.1 Comparison between GSI and stereological data (mean ± SD) from longline- and purse seine-caught bluefin tuna (n: number of individuals examined). When ANCOVA analyses indicated significant differences (p <.5), Dunn-Sidak multiple comparison tests were applied. Values in the same row bearing different superscripts were significantly different. Longline (May-mid June) Longline (mid June-mid July) Purse Seine (mid June-mid July) Males GSI 2.64 ± 1.3 a (n = 6) 3.5 ± 1.81 a,b (n = 22) 4.34 ± 1.68 b (n = 42) Females GSI 2.81 ± 1.39 a (n = 86) 2.86 ± 1.41 a (n = 52) 3.85 ± 1.66 b (n = 56) Lipid-stage oocytes V V Nv (ml -1 ) ( 1 3 ) No per individual ( 1 6 ) No. per g of BW (g -1 ).7 ±.2 b (n = 74) ± 17.2 b (n = 74) ± b (n = 72) ± b (n = 72).5 ±.2 a (n = 45) 2.73 ± a (n = 45) ± a (n = 44) ± a (n = 44).7 ±.4 b (n = 49) ± b (n = 49) ± b (n = 48) ± b (n = 48) Vitellogenic ocytes V V.25 ±.1 (n = 74) Nv (ml -1 ) ( 1 3 ) 9.87 ± 3.65 b (n = 74) No per individual ± ( 1 6 ) (n = 72) No. per g of BW ± (g -1 ) (n = 72).29 ±.13 (n = 45) 9.43 ± 4.71 a,b (n = 45) 51.8 ± 31.5 (n = 44) ± (n = 44).24 ±.11 (n = 49) 7.71 ± 3.68 a (n = 49) 34.4 ± 36.4 (n = 48) ± (n = 48) Final-maturation oocytes V V Nv (ml -1 ) ( 1 3 ) No per individual ( 1 6 ) No. per g of BW (g -1 ). ±.3 a (n = 74).4 ±.21 a (n = 74).2 ± 1.12 a (n = 72) 1.1 ± 5.8 a (n = 72).1 ±.2 a (n = 45).9 ±.27 a (n = 45).45 ± 1.38 a (n = 44) 2.69 ± 8.28 a (n = 44).19 ±.14 b (n = 49) 2.6 ± 2.23 b (n = 49) 9.47 ± b (n = 48) ± b (n = 48) Atretic oocytes V V.6 ±.5 a (n = 74).8 ±.9 b (n = 45).4 ±.6 a (n = 49) Histology Males Histological features of testes from all specimens examined suggested the occurrence of active spermatogenesis in all cases: trap, purse seine- and longline-caught bluefin tuna. At the peripheral region of the testes, the seminiferous lobules show cysts of germ cells at all developmental stages of spermatogenesis (Fig. 3.15A). The central ducts of the testes lack cysts of developing male gametes, and serve for the storage of spermatozoa (Fig. 3.15B). There were no marked histological differences indicative of functional 98

99 differences in testicular activity between longline- and purse seine-caught tuna. Therefore, the overall histological structure of testes can hardly account for the great disparity found in the GSIs of both groups. Females Most females from the trap fishery were classified as active non-spawning (83%), while only 17% of them were inactive; no spawning female was found in migrant tuna caught in the Strait of Gibraltar. Unlike males, female tuna sampled by longline and purse seine differed in overall gonad histology, suggesting a distinct state of maturation. The greater part (96.3%) of the purse-seine sample (52 individuals out of 54 examined for histology) consisted of active spawning females, only one female was active nonspawning and the remaining female was classified as inactive mature (Fig. 3.15A). There were no immature specimens among the female tuna caught by purse seine. In the longline sample 35.6% of the females (47 out of 132) were identified as active spawning, 54.6% were considered active nonspawning, 8.3% inactive mature, and only two individuals (1.5%) were classified as immature (Fig. 3.15A). These proportions change significantly when the longline sample is divided into two fishing periods as indicated above. In the group of females caught during May through mid June, 24.% of the individuals were active spawning and 72.2% active nonspawning, whereas 52.8% of the females caught between mid June and mid July were active spawning and 28.3% active nonspawning (Fig. 3.15A). It is remarkable that the percentage of inactive mature fish was relatively high (17.%) in the sample collected during the second period of the longline fishing season in comparison to earlier catches (2.5%). The low proportion of immature females remained consistent (about 1.5%) throughout the whole longline fishing season. Fig Micrographs of transverse sections of bluefin tuna gonads. (A and B) Testes (H-E staining). (A) Peripheral region showing cysts of developing germ cells. (B) Transition between peripheral (PR) and inner (IR) regions. (C-F) Ovaries of active nonspawning (C), active spawning (D), inactive mature (E), and immature (F) tuna (H-VOF staining). LS: lipid-stage oocytes; MN: migratory nucleus oocytes; n: nucleus; PN: perinucleolar oocytes; POF: postovulatory follicles; sd: late spermatid cysts; sz: free spermatozoa; TA: tunica albuginea; V: vitellogenic oocytes; α: α atretic follicles; β: β atretic follicles. 99

100 Fig Summarized results of the histological analysis in females(longline and purse seine). According to the ovarian histological structure (see text), individuals of each sample group were allocated through four different stages of maturation (Schaefer 1996, 1998, Schaefer et al. 25). The percentages of specimens pertaining to the different histological categories are provided for longline and purse-seine samples (A). Frequencies of females with postovulatory follicles (%) and the associated estimated spawning intervals (days) are also shown (B). In order to provide additional information on histological variations with time in the longline sample, the whole catch (whole) was split into two distinct periods corresponding to early catches (May through mid June, 1st period) (1st) and late catches (from mid June to mid July, 2nd period) (2nd). LL: longline; PS: purse seine. Postovulatory follicles were present in 83.7% of the females caught by purse seine (Fig. 3.16B). Considering a maximum permanence of POFs in the ovary of 24 hours, the average spawning frequency can then be estimated in.84 spawns per day, which corresponds to an average interspawning interval of 1.19 days. If only active tuna are considered, the estimated spawning frequency and spawning interval correspond to.85 day -1 and 1.17 days, respectively. Among all longline-caught fish the specimens containing POFs represented only 32.6% of the sample, which is equivalent to an average spawning interval of 3.1 days. Marked differences appear to occur depending on the date when the samples were taken. Thus, while POFs occurred in only 22.8% of the early catches (between early May and mid June), with a spawning interval of 4.4 days, 49.% of the females caught during the second period of the longline season (from mid June to mid July) had spawned within the previous 24 hours, representing an average interval of 2 days between consecutive spawns (Fig. 3.16B). If only active fish caught during this period are considered, the frequency of ovaries containing POFs is 58.1%, which corresponds to a spawning interval of 1.72 days. This spawning frequency is closer to that found in the purse-seine sample, but yet reflect clear differences between the two sampling gears, even though both captures were carried out in the same period Stereology Table 2 summarizes the results of the stereological analysis in longline and purse-seine samples. For lipidstage oocytes significant differences were found in the values of fractional volume (V V ) and in the estimated number of oocytes per g of BW. In vitellogenic oocytes, a slight but significant difference was present only in the numerical density (N V ), whereas the other parameters showed statistically similar values, especially the relative number of vitellogenic oocytes (number of yolked oocytes per g of BW). The strongest differences between the two samples were encountered in the population of oocytes at final maturation, which includes migratory-nucleus and hydrated oocytes. All the estimates (V V, N V, absolute and relative numbers of oocytes) proved to be statistically different in the two groups (Table 3.9). Particularly, in purse seine-caught tuna the average number of oocytes at final maturation, which may be considered as a reliable proxy of the species batch fecundity, was approximately 9 million (82 oocytes g -1 ), whereas a rather exiguous amount of 3, final-maturation oocytes (less than 2 final-maturation stage oocytes g -1 ) was estimated for the longline sample. 1

101 It should be noted that dehydration of tissues prior to histological embedding results in volume losses that would affect such stereological parameters as N V and its numerical derivatives (i. e., absolute and relative number of oocytes). After applying the previously calculated correction factor of.72, the resulting average batch fecundities are 6.5 million eggs (relative fecundity of 59 eggs g -1 ) in the purse seine and 216, eggs (1.2 eggs g -1 ) in the longline. Considering two separate periods for the longline fishery does not modify greatly these results (Table 3). The amount of final-maturation oocytes present in the gonad were not significantly different between early and late longline catches, both values remaining far lower than in the purse-seine sample. In vitellogenic oocytes differences persisted in N V between early-caught longline tuna and purse-seine tuna, with values in latecaught longline tuna that were intermediate between those of the two other groups, whereas all other stereometric parameters proved statistically similar. In particular, the estimated number of vitellogenic oocytes per g of BW was virtually identical in all three groups (Table 3.9). Interestingly, stereological analysis of lipid-stage oocytes gave similar results in all estimated stereological parameters in early-caught longline tuna and purse-seine tuna, whereas these values were lower in tuna captured by longline at the end of the fishing season (Table 3.9). It appears that the extremely low mean number of final-maturation oocytes estimated for the longline sample in comparison to the tuna captured with purse seine is largely due to the poor proportion of fish containing oocytes of this category in the ovaries. Nevertheless, when only specimens having oocytes at final maturation stage are considered in the statistical analysis, differences persist between both groups, with means of about 95 migratory-nucleus oocytes g -1 in purse-seine tuna versus 21 migratory-nucleus oocytes g - 1 in longline tuna Discussion Results of the trap-caught tuna are very similar between the three years of sampling. Condition indices clearly denote a difference in the state of maturation of breeding tuna with time. As the migratory season advances, the GSI increases while the fat stores (reflected in FBI values) decrease. No significant interannual differences have been observed in these patterns; therefore, this seems to be a general trend in the reproductive biology of the eastern stock of Atlantic bluefin tuna. Marked differences are also evident between the three different gears used for sampling the BFT. The highest GSIs and lowest FBIs corresponded to bluefin tuna captured by purse-seine, whereas the lowest GSIs and highest FBIs were recorded in migrant bluefin tuna caught at the Strait of Gibraltar. Tuna fished by long-line had intermediate values. These observations are consistent with results from histological and stereometric analyses. Our observations on bluefin tuna caught in the Strait of Gibraltar throughout the migratory period in 25 are consistent with those corresponding to years 23 and 24, though differences between early and late samples were less stressed than in the first two years. In general, GSI, FBI and reproductive parameters estimated by stereology proved statistically different between the specimens caught in late April-early May and those fished in late May-early June. Regression analyses of the relationship between GSI and FBI in both females and males indicated a significant negative correlation, which reveals a depletion of mesenteric fat stores as the gonad grows. These observations strengthen the hypothesis that fat body lipid reserves provide an important energy source for gametogenesis in tunas (Medina et al., 22; Abascal et al., 24). Stereometric estimations in females agree with these data, the number of yolked oocytes being much higher in bluefin tuna that are caught at the end of the eastward migration period. It is thought that the increase in ovarian weight is primarily accounted for by an increase in the number of oocytes that have entered active vitellogenesis. In the Mediterranean Sea bluefin tuna are caught mainly by purse seine and longline fleets. It is thought that these gears target fish that display behavioural differences and are probably at different reproductive conditions (Schaefer 21). Male and female bluefin tuna sampled by purse seine around the Balearic Islands had clearly higher GSIs than those caught by longline. However, no remarkable histological differences were found between purse seine- and longline-caught males. Thus, the testis histological structure suggests that males in both samples were ready to spawn. Significantly higher GSIs in mature bluefin tuna males are thought to indicate larger stores of sperm in the central ducts of the testes (Abascal et al. 24). Therefore, in longline-caught males, increased GSIs toward the second part of the fishing season suggest a higher sperm production and storage as the reproductive peak approaches. In females the situation is somewhat different, since the mean GSI appears to remain invariable throughout the entire longline season. Results of the stereological analysis fit well into this data, since the numbers of vitellogenic oocytes and final-maturation oocytes, which are the oocyte classes that most significantly contribute to the gonad weight (Abascal 24), are similar in early and late longline catches. Lipid-stage 11

102 oocytes, however, appear to become more scarce as the longline fishing season proceeds. This might reflect recruitment of lipid-stage oocytes into vitellogenic oocytes to compensate for losses caused by spawning and atresia, thereby maintaining a consistent pool of vitellogenic oocytes throughout the reproductive season. In support of this, the mean estimated number of yolked oocytes per g of BW remained quite constant (between 282 and 285 oocytes g -1 ) irrespective of the time and method of capture. Histological examination of the ovaries revealed indisputable differences in the reproductive condition between early and late longline catches and between both of them and purse-seine catches. For instance, while most females (about 96%) sampled by purse seine were classified as active spawning, only approximately 25% of the females in the early longline catches and 5% in the late longline catches exhibited histological features typical of active spawning fish. Similar percentages were observed for female tuna containing POFs in their ovaries, resulting in estimated spawning intervals of 4.4 days for bluefin tuna caught by longline during May through mid June, 2 days for tuna caught by longline between mid June and mid July, and 1.2 days for those caught by purse seine between mid June and mid July. All the above observations show unequivocally that in the studied fishing ground bluefin tuna targeted by purse seining show a higher proportion of reproductive activity on average than those caught by longlining. This is true even when comparing only the samples collected by the two gear types at the same period, and suggests that differences result basically from the particular nature of the gears themselves. The differences observed in proportions of reproductively active females between the purse seine and longline fisheries are apparently due to differences in behavior and thus vertical distributions of those individuals. Similarly, several authors found a substantially lower reproductive activity in yellowfin tuna caught by longline in comparison to those caught by purse seine in the same area and time (Suzuki et al. 1978, Koido & Suzuki 1989, Bayliff 1994). Davis and Farley (21) showed that size partitioning by depth occurs in the spawning area of the southern bluefin tuna, Thunnus maccoyii. Such a size vertical stratification is apparently related to spawning activity, so that spawning fish seem to be better represented in shallow catches than in deep catches, while nonspawning fish are better represented in deep catches than in shallow catches. It has been hypothesized that size partitioning with depth in this species is due to the fact that large fish spawn more often than smaller ones (Davis & Farley 21). These findings suggest that samples from the longline fishery are not suitable for estimating certain reproductive parameters (e.g., fecundity and spawning frequency). The stereological concept has been applied to ovarian histological sections as an accurate tool for fecundity estimates in fishes (Coward & Bromage 22a). Stereometry is indeed a powerful procedure for quantifying oocytes in three-dimensional samples. However, it has the objection that routine procedures commonly used in histology laboratories may cause significant shrinkage of the tissue samples. We have calculated an average volume loss of 28% during histological processing which should not be ignored when estimating numerical parameters by stereology. Earlier studies based on stereology (e.g., Medina et al. 22) may have reported an overestimated fecundity (93 eggs g -1 ) in not having taken into consideration shrinkage associated with tissue treatment. Thus, the mean batch fecundity of 9 million eggs (82 eggs g -1 ) that is calculated directly from stereometric counts of final-maturation oocytes in purse seine-caught spawners would drop to 6.5 million eggs, equivalent to a relative batch fecundity of 59 eggs g -1, after correction for volume reduction. This value is nearly identical to the relative batch fecundity calculated for southern bluefin tuna, Thunnus maccoyii (Farley & Davis 1998) (57 eggs g -1 ) and is also close to the estimates reported for yellowfin tuna, T. albacares, by Schaefer (1996) (68 eggs g -1 ) and Itano (2) (55 eggs g -1 and 64 eggs g -1 ). Somewhat lower values (24 eggs g -1 ) were obtained for bigeye tuna, Thunnus obesus, by Schaefer et al. (25). Determining the spatiotemporal distribution of bluefin tuna populations is crucial for the resource management. The International Commission for the Conservation of Atlantic Tunas (ICCAT) manages Atlantic bluefin tuna as two separate stock units. Although information provided by electronic tags does support the occurrence of two independent spawning populations, these appear to intermingle in the central Atlantic. A modern view tends to reject the concept of a bluefin tuna stock consisting of geographically and reproductively isolated populations in favour of a metapopulation-like model (Fromentin & Powers, 25). This supports the idea that bluefin tuna function as a set of discrete populations that experience a certain degree of demographic influence from other local populations. Within each local population, individuals would have migratory behaviours and use of common habitats. According to this, global research into bluefin tuna population dynamics would require comparative trans-oceanic surveys using the same sampling protocols (Fromentin & Powers, 25). In particular, reproductive studies should carefully consider the type of gear used in sample collection, since the results obtained may be highly variable depending on the capture methods employed. Bluefin tunas differ from other tuna species in having a reduced spatiotemporal window for reproduction (Schaefer 21). In the West Mediterranean Sea, spawning takes place primarily in June-July, but the time individual fish take to complete the whole series of spawning events is so far unknown. Although bluefin tuna 12

103 breeders are generally believed to reside for about two months in areas of reproduction, archival data suggest that the effective time they spend on spawning grounds might actually be as short as 2 weeks (see Gunn & Block 21, p 212). This is in agreement with the finding that spent Thunnus maccoyii are seldom encountered on their spawning grounds, suggesting that they move south soon after spawning (Farley & Davis 1998). Our results appear to indicate a similar behaviour in T. thynnus, since the low proportions of inactive fish found in both longline and purse-seine catches hint that bluefin tuna leave the spawning grounds shortly after having accomplished their reproductive function, thus becoming invulnerable to those fisheries. Considering the average spawning frequency of.85 day -1 estimated for reproductively active tuna, and the average batch fecundity of 6.5 million eggs, an average individual spawning duration of 14 days in the Balearic spawning ground would yield an average annual fecundity of around 77 million eggs per fish (relative annual fecundity of 72.1 eggs per g of BW). Indeed, an appropriate combination of reproductive biology studies with data generated by currently growing electronic tagging programmes will help obtain a better understanding of the population dynamics of bluefin tuna. Further analysis of movement patterns around reproductive grounds should be encouraged in the near future with a view to improve the assessment of reproduction of Atlantic bluefin tuna stocks and build useful management models. 3.5 DETERMINATION OF STRESS IN WILD BFT POPULATIONS Introduction The three basic criterion for stress used in this study were the levels of Cortisol, the Catecholamines Norepinephrine and Epinephrine, and Lactate in the plasma. The use of these three criteria was chosen to represent chronic and acute stress parameters. Basically cortisol levels may arise within a one to four hour time frame and then remain relatively high, Catecholamine rise relatively rapidly and also lactate levels within 3 minutes Methodology and study material A standard cortisol ELISA will be used. This assay is based on the competition principle and the microtiter plate separation. An unknown amount of antigen present in the sample and a fixed amount of enzyme labelled antigen compete for the binding sites of the antibodies coated onto the wells. After incubation, the wells are washed to stop the competition reaction. Having added the TMB substrate solution the concentration of antigen is inversely proportional to the optical density measured. The measured ODs of the standards are used to construct a calibration curve against which the unknown samples are calculated. Catecholamines: Plasma samples will be extracted using standard procedure with acid activated aluminium powder, washing and then extraction with perchloric acid. Sample will then be measured with HPLC and electrochemical detection. External standards for Epinephrine and Nor-epinephrin will be used together the internal standard DHBA to control extraction efficiency. Lactate was assayed with a standard optical test and standards Results Season 23 Cortisol Measurements From figure 3.18, four different sampling areas within Italian waters are defined: 1. The Gulf of Taranto Drift net 2. The Straits of Messina Hand line 3. The Southern Tyrrhenian Sea (Sicily) Purse seiner 4. Sardinian Channel Mattanza Trap 13

104 Plasma Cortisol in Wild Fish (Italy) Gulf of Taranto Straits of Messina South Tyrrhenian Sea (Sicily) Sardinia Channel Cortisol Conc. (ng/ml) B 1 B 2 B 3 B 4 B 5 B 6 B 7 B 8 B 9 B 1 B 11 B 12 B 13 B 14 B 15 B 16 B 17 B 18 B 19 B 11 B 111 B 113 B 114 B 115 B 116 B 117 B 118 B 119 B 12 B 121 B 122 B 123 B 124 B 125 B 126 B 127 B 128 B 129 B 13 B 131 B 132 B 133 B 134 B 135 B 2 B 21 B 22 B 23 B 24 B 25 B 26 B 28 B 29 B 21 B 211 B 212 Fish Number Fig Plasma Cortisol values in four different Italian sampling areas. It is clear that fish caught in the Straits of Messina had plasma cortisol values four times higher (mean value =228ng/ml) than those in the Gulf of Taranto or the southern Tyrrhenian Sea (58ng/ml). Whereas in general the values in the Gulf of Taranto and the Southern Tyrrhenian Sea averaged below 6 ng/ml, those in the Sardinian Channel were around 9ng/m Table 3.11 Mean and STD for Cortisol Sampling Area Mean Cortisol Conc. (ng/ml) STD Mazarron 146,5 87,4 Bari-Taranto 59,3 51,1 Messina 228, 78,2 Tyrrhenian 52,5 27,7 Sardinia 89,8 67,6 Cadiz Almadrava Zahara 68,1 41,9 Almadraba Conil 17,5 36,3 Fuengirola Balearic Island 192,5 136,8 Table 3.11 and Figure 3.19 give the details for the Cadiz sampling area at two different tuna traps. The cortisol levels are relatively low. The capturing methods and killing methods must be looked at in more detail to determine the reasons for any differences. 14

105 Plasma Cortisol in Wild Fish (Cadiz) Almadraba de Zahara Almadraba de Conil Cortisol (ng/ml) C1 C4 C6 C8 C1 C12 C14 C18 C2 C22 C24 C26 C28 Fish Number Fig Plasma cortisol values for wild fish from the tuna traps in the Cadiz area In Figure 3.2 and table 3.11 the levels for the Balearic fish are shown. These fish caught by purse-seiners showed the highest cortisol levels in terms of the mean values. Plasma Cortisol Levels in Balearic Fish Cortsiol Conc. (ng/ml) F 17 F 23 F 29 F 33 F 36 F 39 F 42 F 46 F 47 F 48 F 51 F 53 F 54 F 55 F 13 F 2 F 21 F 22 F 27 F 41 F 44 F 5 Fish Number Fig. 3.2 Plasma cortisol levels in wild fish for the Balearic Island Area in 23 Catecholamine Values During the latter sampling sessions in June and July extremely high values were measured indicating that acute stress had occurred through these handling techniques. For comparative purposes catecholamine levels were measured in wild fish from Italian sampling areas and also from the Cadiz sampling areas. These are shown in Figures 3.21 and 3.22, and again indicate the general correlation between cortisol and catecholamines; however there are specific differences as can be seen with the Straits of Messina fish. In these fish plasma Norepinephrine an Epinephrine levels were extremely low. Epinephrine appears to accumulate too much higher levels than Norepinephrine. In the Cadiz fish most of the epinephrine values were above 5 nm which appears to be roughly the threshold beyond which acute stress appears to have occurred. 15

106 Plasma Nor- and Epinephrine Wild Fish (Italy) 23 6 Gulf of Taranto Straits of M essina South Thyrrhenian Sea (Sicily) Sardinia Channel 5 Catechol Conc. (nm) Norepi Epi 1 n.d B 1 B 2 B 3 B 4 B 5 B 6 B 7 B 8 B 9 B 1 B 11 B 12 B 13 B 14 B 15 B 16 B 17 B 18 B 19 B 11 B 111 B 113 B 114 B 115 B 116 B 117 B 118 B 119 B 12 B 121 B 122 B 123 B 124 B 125 B 126 B 127 B 128 B 129 B 13 B 131 B 132 B 133 B 134 B 135 B 2 B 21 B 22 B 23 B 24 B 25 B 26 B 28 B 29 B 21 B 211 B 212 Fish Number Figure Plasma catecholamines in Italian wild fish from 4 different sampling sites and using different techniques Plasma Nor- and Epinephrine in Wild Fish (Cadiz) Almadraba de Zahara Almadraba de Catechols (nm) C1 C3 C4 C5 C6 C7 C8 C9 C1 C11 C12 C13 C14 Fish Number Figure 3.22 Plasma catecholamines in wild fish from the traps in the Cadiz area. Nor- Epi C17 C18 C19 C2 C21 C22 C23 C24 C25 C26 C27 C28 C29 It is evident that there is a correlation between plasma cortisol and the catecholamines; however they both may have different time frames of release. The general trend is of course that with increasing cortisol levels Catecholamine levels also increase and the correlation although is poor and again epinephrine values rise to much higher values than those for Norepinephrine. The general trends are clear however and it should be possible to develop scales of stress levels defined by plasma Lactate cortisol and catecholamine levels. Further trial experiments were made with measurements of mucus levels of cortisol, steroids and vitellogenin (see later report). However there were not enough samples to make any plausible analysis possible. It was noted that cortisol could be detected in mucus. The relative a threshold values for the three per determinants in plasma, cortisol, catecholamines and lactate a highly dependent upon the levels of stress induced and whether or not acute or chronic stress is to be measured. In this case then only relative levels can be used and not absolute levels. For Cortisol plasma levels below 5 ng/ml represent a lower state of stress. Between 5 and 1 ng/ml stress factor increases 16

107 and values above 1 appeared to be indicative of chronic stress. The Mazarron caged animals over the four sampling periods had a relatively high cortisol values indicating perhaps the state of chronic stress. It is obvious in wild populations that lower levels can be found. From the 4 different sampling methods used in the Italian fishing campaign it could be noted that the purse seiners had cortisol levels below 5 ng/ml whereas those caught on a hand line were extremely high. The type of capture technique used and the interval between the onset of stress and the actual fish being sacrificed are also variable. These will affect not only short term indicators such as lactate and catecholamines, but also the long-term release of cortisol. In general confinement as in the Mazarron cages, or in the Cadiz sampling areas or in Sardinia lead to raise cortisol values. Catching up by a baited hook as in the first and second sampling periods in Mazarron and also in the Straits of Messina lead to much lower catecholamine levels in the plasma, emphasising the need for the measurements are both cortisol and plasma catecholamines. Cortisol and Lactate Comparison For the wild sampling season 23 the measurements of Lactate and Catecholamines were completed for samples from Bari (diverse sampling areas), Barbate (Cadiz) and Balearic (Malaga). The figure below (fig. 3.23) compares the results reported last year for Cortisol in Barbate with the lactate levels. Whereas Cortisol values ranged from 1 ng/ml to 17 ng/ml lactate levels were almost constant in the plasma at 2 mm. Plasma Cortisol in Wild Fish (Cadiz) 2 18 Almadraba de Zahara Trap Almadraba de Conil Trap Cortisol (ng/ml) and Lactate (mm) Cortisol Lactate C1 C3 C4 C5 C6 C7 C8 C9 C1 C11 C12 C13 C14 C17 C18 C19 C2 C21 C22 C23 C24 C25 C26 C27 C28 C29 Fish Number Fig Plasma cortisol and lactate levels in Tuna caught in traps in 24 From the measurements of lactate from the Italian sampling areas there is a good correlation between cortisol and lactate to some extent (See Fig 3.24). Very high values for lactate approaching 5mM were observed in some fish. A level that would almost certainly cause death and can be attributed to the catching methods used. Again as with cortisol lower lactate values were found in the South Tyrrhenian Sea areas and also the Sardinia channel. 17

108 Plasma Cortisol and Lactate in Wild Fish (Italy) Cortisol Conc. (ng/ml) and Lactate (mm) Gulf of Taranto Drift net Straits of Messina Hand-Line South Thyrrhenian Sea (Sicily) Purse Seiner Sardinia Channel Trap Cortisol Lactate B 1 B 2 B 3 B 4 B 5 B 6 B 7 B 8 B 9 B 1 B 11 B 12 B 13 B 14 B 15 B 16 B 17 B 18 B 19 B 11 B 111 B 113 B 114 B 115 B 116 B 117 B 118 B 119 B 12 B 121 B 122 B 123 B 124 B 125 B 126 B 127 B 128 B 129 B 13 B 131 B 132 B 133 B 134 B 135 B 2 B 21 B 22 B 23 B 24 B 25 B 26 B 28 B 29 B 21 B 211 B 212 Fish Number Fig Plasma lactate and cortisol levels in wild fish caught in Italian waters in 23 Cortisol values shown for the Balearic fish have been complimented with the lactate levels as shown below (Fig. 3.25). In general from these fish, Cortisol levels were high together with Lactate in fish which were caught by long-line. Lactate levels varied between 8 and 53 mm and Cortisol below resting values of 1 ng/ml and 54 ng/ml. Plasma Cortisol and Lactate Levels in Balearic Fish Cortsiol Conc. (ng/ml) & Lactate (mm) F 13 F 17 F 2 F 21 F 22 F 23 F 27 F 29 F 33 F 36 F 39 F 41 F 42 F 44 F 46 F 47 F 48 F 5 F 51 F 53 F 54 F 55 Fish Number Cortisol Lactate Fig 3.25 Cortisol and lactate values for plasma from wild fish caught around the Balearic Islands in 23. Although cortisol was high the values for the Catecholamines Norepinephrine and Epinephrine were relatively low below 2nM in many fish. Again the catecholamine levels may be a function of the capture method used. 18

109 Plasma Catecholamine Levels in Balearic Fish 23 Plasma Catecholamines (nm) F 13 F 17 F 2 F 21 F 22 F 23 F 27 F 29 F 33 F 36 F 39 F 41 F 42 F 44 F 46 F 47 F 48 F 5 F 51 F 53 F 54 F 55 Fish Number Norepine Epinep Figure 3.26 Plasma Catecholamines for the Balearic fish in Season 24 In the season 24 sampling was carried out in captive fish in Mazarron and these formed the basis of most of the measurements to date. At the same time samples were collected from wild fish in the Balearic Islands, In the Barbate traps and also from various areas around Italy. From the relative few results from the catecholamine determinations made so far samples have been measured for the central Mediterranean and the Barbate traps as shown below in the two figures (Fig 3.27 and Fig 3.28). Plasma Catechols 24 in Central Mediterranean Malta North Ionian Sea 4474 Sardinia Trap Catechol Conc. (nm) Norepinephrine Epinephrine 2 B41 B42 B43 B44 B45 B46 B47 B48 B49 B41 B411 B412 B413 B414 B415 B42 B421 B422 B423 B424 B425 B426 B427 B428 B429 B43 B431 B432 B433 B434 B435 B436 B437 B438 B439 B44 B441 B442 B443 B444 B445 B446 B447 B448 B449 B45 B451 Fish Nr. Fig Catecholamine values in the plasma of Central Mediterranean Fish in 24 The variability in the central Mediterranean samples is probably linked with the method of capture. 19

110 Plasma Catechols Barbate Almadraba de Barbate Almadraba de Zahara 4 Catechols (nm) Norepin Epineph C35 C36 C37 C38 C39 C4 C41 C42 C43 C44 C45 C46 C47 C48 C49 C5 C51 C52 C53 C54 C55 Fish Nr. Fig Catecholamine values in the plasma of Barbate Trap Fish in Discussion-Conclusions From analysis of the wild fish in 23 for lactate it is clear that the levels experienced by many of the fish are lethal over an acute exposure. The long-term patterns of cortisol, catecholamines and lactate neeed to be analysed using all the data. It is evident from both the lactate and catecholamine levels that the caged fish are stressed at capture. There was a clear increase with time in lactate levels as the fish were slaughtered. This was in reverse to the observations made for the catecholamines. An in depth study of the data so far will be made when time permits. From the few measurements made in tissues these show much lower levels as expected for cortisol and catechols. From the 24 sampling it was possible to establish an excellent linear correlation between tissue and plasma lactate levels. This may be used in future in tagging techniques to assess stress levels at tagging during sex determination measurements. Tissue lactate is a good indicator of acute stress and may be a valuable indicator for future aquaculture work Wild Fish Barbate 24 and 25 As previously observed Norepinephrine values were lower than the Epinephrine values in all the fish measured (Fig. 3.29). The catecholamine values in both 24 and 25 were relatively high indicating that the fish were already stressed by the slaughtering procedure. Lactate values in the 24 fish were also already high at around 2 mm (Fig. 3.29). The cortisol values in the 24 fish (Fig. 3.29) indicated that chronic stress had taken place whereas in 25 cortisol values are much reduced in a number of fish to values below 7 ng/ml (Fid. 3.3). This will very much depend on the operation of the trap and also the slaughtering procedures. 11

111 Barbate Traps (Cadiz) 24 Catecholamines C35 C36 C37 C38 C39 C4 C41 C42 C43 C44 C45 C46 C48 C49 C5 C51 C53 C54 C55 Catecholamine concentration [nmol/ml] Fish zz number Norepinephrin Epinephrin 3 Barbate Traps (Cadiz) 24 Lactate and Cortisol Lactate concentration [µmol/ml] C35 C36 C37 C38 C39 C4 C41 C42 C43 C44 C45 C46 C48 C49 C5 C51 Cortisol concentration [ng/ml] C53 C54 C55 Fish number Lactate Cortisol Fig The plasma levels for catecholamines, lactate and cortisol of fish caught in the Barbate traps are shown for 24. Fish numbers are shown at the bottom. 111

112 Catecholamin concentration [nmol/l] Barbate Traps 25 Catecholamines + Cortisol C61 C62 C63 C65 C66 C67 C68 C69 C7 Cortisol (ng/ml) Fish number Norepinephrin Epinephrin Cortisol Figure 3.3 The combined results for catecholamines and cortisol in the plasma fish caught in Barbate trap in Balearic Islands (24) In the Balearic Islands fish are normally caught using long lines. Depending upon the time spent on the hook and the time before hauling the line, the values for the stress indicators will be affected accordingly. This is shown in the following figures (Fig. 3.31) as in some fish high levels of catecholamines are reached whereas in others catecholamines are low or not present. The general trend in 24 was that the values in May were somewhat lower than those in June, with a decrease again in July. It is unknown whether this has any correlation with the spawning season or the general levels of activity of the fish themselves. It is also to be noted that some of the fish were alive when the longline was hauled. A more detailed analysis of the data will be carried out looking for correlations between values in live fish when the data is prepared for publication. 112

113 Balearic Islands (Malaga) 24 Catecholamines - Part I 2 May June F-1 F-2 F-3 F-4 F-5 F-6 F-7 F-8 F-9 F-21 F-1 F-11 F-12 F-22 F-13 F-14 F-15 F-16 F-17 F-18 F-19 F-2 F-23 F-25 F-31 F-32 F-26 F-27 F-28 F-33 F-34 F-35 F-36 Catecholamine concentration [nmol/ml] F-37 F-29 Fish zz number Norepinephrin Epinephrin Balearic Islands (Malaga) 24 Catecholamines - Part II 18 June July Catecholamine concentration [nmol/ml] F-3 F-38 F-39 F-4 F-41 F-42 F-43 F-44 F-45 F-46 F-47 F-48 F-49 F-5 F-51 F-52 F-53 F-54 F-55 F-56 F-57 F-58 F-59 F-6 F-61 F-62 F-63 F-64 F-65 F-66 F-67 F-69 F-7 F-71 Fish zz number Norepinephrin Epinephrin Figure 3.31 Plasma catecholamine values for fish caught in 24 in the months of May, June and July around the Balearic Islands by long lines The trends in plasma lactate and cortisol shown in figure 3.32 indicate predominantly that an acute stress was the main factor in the longline fish caught in 24. In numerous fish values for plasma lactate exceeded 5 mm and very few values were under 2 mm. Cortisol values however were relatively low and did not exceed a 1 ng/ml. This compares with values well over 1 ng/ml in the Barbate trap fish. It would therefore appear that cortisol levels as expected rise slowly during chronic exposure, whereas the time period for lactate and catecholamines is short. 113

114 6 May Balearic Islands (Malaga) 24 Lactate and Cortisol - Part I June 2 Lactate concentration [µmol/ml] Cortisol concentration [ng/ml] F-1 F-2 F-3 F-4 F-5 F-6 F-7 F-8 F-9 F-21 F-1 F-11 F-12 F-22 F-13 F-14 F-15 F-16 F-17 F-18 F-19 F-2 F-23 F-25 F-31 F-32 F-26 F-27 F-28 F-33 F-34 F-35 F-36 F-37 F-29 Fish number Lactate Cortisol Balearic Islands (Malaga) 24 Lactate and Cortisol - Part II June July Lactate concentration [µmol/ml] Cortisol concentration [ng/ml] F-3 F-38 F-39 F-4 F-41 F-42 F-43 F-44 F-45 F-46 F-47 F-48 F-49 F-5 F-51 F-52 F-53 F-54 F-55 F-56 F-57 F-58 F-59 F-6 F-61 F-62 F-63 F-64 F-65 F-66 F-67 F-69 F-7 F-71 Fish number Lactate Cortisol Fig Plasma lactate and cortisol values for fish caught in May, June and July 24 by longline around the Balearic Islands Central Mediterranean (24) In 24 plasma samples were obtained from the Sardinian traps, long lines from Malta and gill nets from the North Ionian Sea. In the trap, animals catecholamine levels were extremely low or non-existent apart from fish slaughtered at the end of the sampling period (see fig. 3.33). This correlates well with the relatively low cortisol values at the beginning of the sampling period which gradually increase as sampling continues. The lactate levels in the trap fish averaged around 2 mm. In the Malta longline fish catecholamine levels were low, and unfortunately the cortisol and lactate levels could not be determined (Fig. 3.33). The gill net fish from the North Ionian Sea showed low catecholamine levels, extremely low cortisol levels and high lactate levels. 114

115 Catecholamine concentration [nmol/ml] 14 Sardinian Traps Central Mediterranean 24 Catecholamines Malta North Ionian B42 B421 B422 B423 B424 B425 B426 B427 B428B429 B43 B431 B432 B433 B434 B435 B436B437B438B439B44B441B442B443B444B445B446B447B448B449B45B451B41B42B43B44B45B46B47B48B49 B411B412 B413 B414 B415 Fish zz number Norepinephrin Epinephrin 6 Sardinian Traps Central Mediterranean 24 Lactate and Cortisol Malta North Ionian B42 B421 B422 B423 B424 B425 B426 B427 B428 B429 B43 B431 B432 B433 B434 B435 B436 B437 B438 B439 B44 B441 B442 B443 B444 B445 B446 B447 B448 B449 B45 B451 B41 B42 B43 B44 B45 B46 B47 B48 B49 B411 B412 B413 B414 B415 Lactate concentration [µmol/ml] Cortisol concentration [ng/ml] n.d. Fish number Lactate Cortisol Figure 3.33 Plasma catecholamine, lactate and cortisol levels in fish from three different areas of the central Mediterranean. The Sardinian fish were caught in traps; the Malta fish by longline; and the North Ionian fish by gill nets. The data is shown in chronological order. It appears that there is a general correlation between method of capture and the level of the stress indicators in the central Mediterranean fish with lactate as the acute response followed by slower increases in catecholamines and cortisol Balearic Islands (25) The fish sampled in the Balearic Islands in 25 were mostly caught by longline and showed similar trends to those in 24 (Fig. 3.34). A seasonal trend with low values in May followed by a higher values in June and then low values in July could also be discerned although there were some exceptions which probably could be traced to different catching times. Cortisol values also followed a similar trend with values dropping below 1 ng/ml in July. This tendency was also seen in the sub sample made from a purse seine from the Cadiz group (data not shown here). There are again exceptions but a deeper cause and analysis are required to find the reasons for these variations. 115

116 Catecholamine concentration [nmol/l] Balearic Islands (Malaga) 25 Catecholamines + Cortisol - Part I May F - 16 F F F F F F F F F F F F F - 17 F F F F F F - 27 F F F F F F F F F F F F F F F F F F F F Fish number June Cortisol (ng/ml) Norepinephrin Epinephrin Cortisol Catecholamine concentration [nmol/l] Balearic Islands (Malaga) 25 Catecholamines + Cortisol - Part II June July F F - 21 F - 23 F - 24 F F - 27 F F F F F - 29 F F F F F F F F F F - 3 F - 46 F F F F Cortisol (ng/ml) F F F F F F - 47 F F Fish number Norepinephrin Epinephrin Cortisol Fig Combined results for plasma catecholamine and cortisol levels for longline fish caught around the Balearic Islands in May, June and July 25. Fish number is given below in the diagram. They appeared to be little correlation between catecholamines and lactate or cortisol. Lactate levels also in the fish were low when compared with wild fish populations. This would again indicate that the slaughtering technique of shooting the tuna through the head causes less stress than found in either long lines, purse seining or gill nets. The highest lactate value in this series experiments was 16 mm compared to 33 mm in 24 and over 75% of the cortisol values in 25 were below 2 ng/ml compared to less than 5% in 24. In the three fish a with values above 5 ng/ml no indication of contributory acute stress factors such as a long chasing period, prolonged time period before death could be correlated with these values. 116

117 3.6 DETERMINATION OF PLASMA STEROIDS, VTG and ZRP IN WILD BFT POPULATIONS IN COMPARISON WITH CAGED POPULATIONS Introduction Within the framework of a previous study it has been possible to develop methodologies for the measurement of the concentrations of the steroid hormones, Estradiol (E2), 11-Ketotestosterone (11-KT), Testosterone (T) and more recently sulphated 17,2 beta-dihydroxy-4-pregnen-3-one (17,2 beta-p). In a similar manner the methodology for Vitellogenin (Vtg) and Zona Radiata Protein (ZRP) have also been studied and together a diagnostic suite of ELISA s put together for use in tuna research. This has been at some considerable expense in both financing and time to provide the necessary antibodies. The changes in the concentration levels of steroids are good indicators of the reproductive state of fish during the reproductive cycle. The role of 172 beta-p, as a key steroid which is released just before spawning into the plasma, and is rapidly degradated afterwards, may be used as a relative marker for spawning itself. Similar roles in terms of biomarkers are played by vitellogenin and ZRP, which are again correlated with the maturation of the gonads and our indicators of the progression of the reproductive cycle. In the REPRODOTT project both wild fish and caged fish were sampled. The final report presented here considers samples of plasma taken from wild fish and caged fish only. Further studies on muscle samples were not carried out due to time and financial constraints Materials and Methods Steroids Steroid Hormone ELISA Methodology: I. Preparation for measurement - Plasma of animals, stored at 2 C, was thawed and the steroids extracted twice with dichloromethane, the organic phase evaporated. The residue solubilized in the appropriate buffer for ELISA processing (Cuisset et al., 1994) II. Quantitative measurements - for detection and measurement of both 17ß-estradiol, 17,2ß-P, Testosterone and 11- Keto-testosterone used standard ELISA procedures elaborated for other fishes (Cuisset et al., 1994) and presently in routine use in the laboratory Vtg and ZRP Vitellogenin determination in plasma Standard methods has previously derived in our laboratory (Susca et al., 21). Were used for all measurements; these consist of the following steps: I. Identification and purification. Plasma of female animals (with high gonado-somatic index), stored at 2 C were centrifuged and applied to a Biogel column and then an ion-exchanger (Resouce Q). Adsorbed proteins were eluted with a linear gradient of NaCl. Absorbance of the eluted fractions was measured at 28 nm. Eluted fractions containing vitellogenin were identified on SDS-PAGE and then concentrated using an Amicon cell to the desired protein concentration. The whole procedure was performed at 4 C and the protein concentration determined by Bradford method. Molecular weight determination was carried out both by gel filtration on Superose-6, and by SDS electrophoresis II. Immunological procedures -Antibodies. The vitellogenin preparation was mixed with complete Freund s adjuvant and injected subcutaneously in rabbits. After the immunisation procedures the serum was harvested (approximately 8 weeks). To remove the antibodies which react to common serum proteins, anti-vitellogenin was absorbed overnight at 4 C with male plasma. After centrifugation, this antibody was stored at 2 C in glycerol (Bon et al., 1997) Western Blotting. Immunoblotting was used to identify the vitellogenin and it s specificity in plasma samples (Kishida et al., 1992; Mañanos et al., 1994a, b; Bon et al., 1997). III. Quantitative measurements with ELISA. The immunoassay for BFT vitellogenin has been developed (Brdiges et al 21) and validated as for other fishes (Nuñez-Rodriguez et al., 1989; Kishida et al., 1992; Bon et al., 1997) and is based on ELISA developed for salmonid and cyprinid fish and tested using plasma. This system has been developed from previous EU projects for the characterisation of the reproductive cycle of tuna (FTMED 97/ 29). The antibody used in this study it is a polyclonal rabbit anti-tuna Vtg Antibody which has been immuno-precipitated with plasma from a male tuna. 117

118 ZRP determination in plasma Briefly using material from a ripened gonad of the tuna ZRP was extracted according to Oppen-Berntsen et al. (199). The extracted ZRP material was then quantified followed by a molecular weight determination. This was then used as above to generate a specific polyclonal anti- ZRP antibody. Qualitative measurements using an indirect competitive ELISA were then carried out similar to the work on Vitellogenin to determine concentrations in the plasma Results Steroids Wild fish and Cage Fish 23 Sex Steroids Conc. (ng/ml) Gulf of Taranto Strait of Messina South Thyrrhenian (Sicily) Sardinia Channel B 1 B 2 B 3 B 4 B 5 B 6 B 7 B 8 B 9 B 1 B 11 B 12 B 13 B 14 B 15 B 16 B 17 B 18 B 19 B 11 B 111 B 113 B 114 B 115 B 116 B 117 B 118 B 119 B 12 B 121 B 122 B 123 B 124 B 125 B 126 B 127 B 128 B 129 B 13 B 131 B 132 B 133 B 134 B 135 B 2 B 21 B 22 B 23 B 24 B 25 B 26 B 28 B 29 B 21 B 211 B 212 Fish Nr. 11_KT Estradiol Testosterone Figure 3.35 Plasma steroid hormone values measured for wild fish from 4 Italian sampling sites caught between and As would be expected the estradiol levels were higher in the females than in the male fish as shown in Figure 3.35 for all the Italians sampling areas. As the spawning season approached the values were higher in the South Tyrrhenian Sea and the Straits of Messina and lower in Sardinia Channel and the Gulf of Taranto 118

119 Almadraba de Zahara Almadraba de Conil Steroid Conc. (ng/ml) KT Estradiol Testosterone 2 1 C1 C3 C4 C5 C6 C7 C8 C9 C1 C11 C12 C13 C14 C17 C18 C19 C2 C21 C22 C23 C24 C25 C26 C27 C28 C29 Fish Nr. Figure 3.36 Plasma steroid hormone values for wild fish from fish traps near Cadiz caught on the and respectively. From the above figure which shows steroid levels in fish caught in the traps in Barbate. in April and June at two separate traps. It is clear that testosterone levels especially have risen in June but no marked changes in the other steroids can be detected KT Estradiol Testosterone Steroid Conc. (ng/ml) BFT 1 BFT 2 BFT 8 BFT 11 BFT 14 BFT 15 BFT 16 BFT 21 BFT 23 BFT 28 BFT 32 BFT 39 BFT 41 BFT 42 BFT 44 BFT 48 BFT 5 BFT 51 BFT 53 BFT 54 BFT 56 BFT 57 BFT 58 Fish Nr. Figure 3.37 Plasma testosterone and estradiol values for wild fish from around the Balearic Islands caught between and From figure 3.37 it would appear that the steroid levels here peak in the end of May beginning of June with a marked decrease in fish caught after the 18 th of June. 119

120 KT Estradiol Testosterone Steroid Conc. (ng/ml) Z1a Z2 Z2a Z3 Z3a Z4 Z11 Z31 Z32 Z33 Z34 Z35 Z36 Z37 Z38 Z39 Z4 Z41 Z42 Z43 Z44 Z46 Z47 Z49 Z5 Z51 Z52 Z53 Z54 Z55 Z56 Fish Nr. Figure 3.38 Plasma steroid hormone values for female and male fish during the 23 sampling season in Mazarron cage fish ( ) In general and as expected all the male fish showed low levels of estradiol compared to female fish but it was interesting to note that the vales were not zero (figure 3.38). In the February samples the values in the male fish were all low as would be expected in non spawning fish. The May and June samples showed the highest levels of both estradiol and testosterone as would be expected in fish preparing for spawning. Estradiol levels rose to a maximum around 12ng/ml whereas at the same time Testosterone rose in both males and females. In July sex steroid levels had already begun to fall back to background levels as seen in February. This seems to indicate that hormonal side of sexual induction was not disturbed in the Mazarron fish but gonad development was limited. Wild fish and Cage Fish 24 In 24 the fish caught in the central Mediterranean came from Sardinia caught by trap in May and June and from the North Ionian Sea in July and September caught by gill netting. In almost all the fish caught in the central Mediterranean the testosterone values were the highest. This contrasts with the Barbate data from the Western Mediterranean, however this difference may be explained by the difference in the seasonal sampling. In Male fish in general the 11-KT values always dominated and were relatively absent in female fish. In May the levels in the Sardinian channel were relatively low in all fish, an increase was observed in June followed by a decrease in July. This is to be expected as the spawning season is normally in June. The Barbate fish sampled in early May showed lower steroid levels than the ones sampled at the end of May. In both sampling sites estradiol levels rose as the month of June was reached. 12

121 Bari 24 Sex Steroids May (Sardinia) June (Sardinia, Malta ) July (North Ionian) September (North Ionian) Steroid Conc. [ng/ml] KT Estradiol Testosterone 4 2 F F M M M F F M M M F M M M M M F F F Figure 3.39 Plasma sex steroid data for Female(F) and Male (M) Fish from the central Mediterranean in differing regions and seasons. Sex Steroids Barbate 24 9 Almadraba de Barbate (7-5-24) Almadraba de Zahara ( ) Steroid Conc. [ng/ml] KT Estradiol Testosterone 2 1 F F F M F F M F M F F F M F F M M Figure 3.4 Similar data drawn for sampling carried out from the traps at Barbate (Spain) in the beginning and end of May 24 Wild fish and Cage Fish 25 No plasma samples were obtained from the Central Mediterranean for the fishing season 25 due to the decision of the partner not to sample for plasma. 121

122 Barbate Sex Steroids Barbate 25 (3-5-25) 7 6 Steroids [ng/ml] KT Estradiol Testosterone 1 F F F F F F M M M Figure 3.41 Plasma sex steroid levels for Barbate trap fish (Western Mediterranean) in 25. From the details shown in Fig the steroid levels again in Barbate in 25 were similar to those shown in 23 and 24. In the female's testosterone was the dominant steroid and in the males 11-KT, which was relatively absent from the females. Again as sampling was carried out in early May the levels were relatively low. 17,2ßP Barbate 25 (3-5-25) Steroid Conc. [ng/ml] F F F F F F M M M Figure 3.41 Plasma concentrations of the 17,2 β-p In Barbate fish. Captions as above. The plasma steroid 17,2 β-p, which can be used as an indicator of ovulation and it is clear that only background levels of this precursor are present in the Barbate animals as shown in Fig

123 Balearic Islands Sex Steroids Balearen (Cadiz) 25 (2-7-25) Steroid Conc. [ng/ml] KT Estradiol Testosterone 2 F F F M M M M M Figure 3.42 Plasma sex steroid levels in Balearic fish caught using a purse seine at the beginning of July. Sex Steroids Baleares 25 (Malaga) Females 12 Steroid Conc. [ng/ml] May June July 11-KT Estradiol Testosterone F F F F F F F F - 27 F F F F F F F F F F F F F F F F - 21 F - 24 F F - 27 F F F F F F F - 3 F F F F F F F F F F Sex Steroids Baleares (Malaga) 25 Males 16 May June July 14 Steroid Conc. [ng/ml] F - 16 F F F F F F F F - 17 F F F F F F F F F F - 23 F F F - 29 F F F F F F - 46 F F KT Estradiol Testosterone Figure 3.43 Plasma steroids in female and male fish caught by long-line around the Balearic Islands in May, June and July. The numbers (F-xxx) below on the graph referred to the fish numbers allocated to each area. 123

124 A full sampling program was carried out near the Balearic Islands by both the Cadiz group (Fig. 3.42) and the Málaga group (long-liner), although the Cadiz groups sampling was limited to one days sampling from a purse seiner. The dominance of 11-KT in the male fish was again apparent and in general levels in June were higher than in May or July (Fig. 3.43). A peak for 11-KT was also seen in June, and in late July and sex steroid levels in males decreased appreciably. In general the concentrations of sex steroids in males were always higher than the females. 17,2ßP Baleares (Malaga) 25 Females May June July Steroid Conc. [ng/ml] F F F F F F F F - 27 F F F F F F F F F F F F F F F F - 21 F - 24 F F - 27 F F F F F F F - 3 F F F F F F F F F F ,2ßP Baleares (Malaga) 25 Males 9 8 May June July F - 16 F F F F F F F F - 17 F F F F F F F F F F - 23 F F F - 29 F F F F F F - 46 F F - 47 Steroid Conc. [ng/ml] Figure 3.44 Plasma concentrations of 17,2 β-p in female and male fish from around the Balearic Islands in May, June and July. The levels of the ovulation or spawning indicator 17,2 β-p remained at a background of around 2 to ng/ ml (Fig. 3.44). In the month of June some fish did however show higher values between 9 and 12 ng/ml. Since this steroid tends to increase just before ovulation (1 days) and then decrease rapidly after spawning it is difficult to identify peaks of activity. It is however clear that some of the fish sampled had either spawned all were preparing to spawn. 124

125 Mazarron Captive Fish 25 A similar type of analysis was carried out for the sex steroids in the plasma of the Mazarron captive fish, which were either control fish or induced fish. The general trend of high testosterone concentrations in females, accompanied by low 11-KT values and higher 11-KT values in males was again observed. Estradiol values were higher in the females which should have been mature and ready to spawn. The role of the gonadotropin releasing hormone in inducing spawning is dealt with elsewhere, together with its application. Sex Steroids Mazarrón 25 Females Steroid Conc. [ng/ml] treated 11-KT Estradiol Testosterone Sex Steroids Mazarrón 25 Males Steroid Conc. [ng/ml] treated 11-KT Estradiol Testosterone Figure 3.45 Plasma concentrations of sex steroids measured in control and induced Fish in the Mazarron cages in 25. From the result shown in Fig there was no great difference in the estradiol levels between males and females either treated or untreated. The plasma concentrations of 11-KT were approximately 5 times higher in the treated males compared with the treated and untreated females. In both treated males and females testosterone levels were relatively similar and both showed a tendency to increase in the treated fish, although this was not statistically significant with the small sample size used. The plasma concentration of 17,2 β-p increase from untreated to treated fish at least for the mean values and for some of the control 125

126 fish, and especially one male, high levels of 17,2 β-p were observed (Fig. 3.46). Overall in comparison with the wild fish see previous pages, the values for all sex steroids were reduced. Mazarron Females ,2 Beta-P (ng/ml) Control Control Control Control Control Control Control Treated Treated Treated Treated Treated Treated Treated Treated Treated Treated Mazarron Males ,2 Beta-P (ng/ml) Control Control Control Control Control Control Control Treated Treated Treated Treated Treated Treated Treated Treated Treated Treated Treated Treated Treated Treated Treated Treated Figure 3.46 Plasma concentrations of 17,2 β-p in female and male control or treated fish from Mazarron

127 Vitellogenin and ZRP Levels in Plasma Correlation between Plasma Vtg and Fork Length In Fig the relationship between plasma her vitellogenin concentration (mg/ ml) and the length of female fish (cm) are shown for the various sampling areas in Females 1 Mazzaron Bari Barbate Balearen Vtg Plasma (mg/ml) Figure 3.47 Relationship between plasma Vtg levels (mg ml -1 ) and fork length (cm) in female tuna from different sampling areas. Vitellogenin concentrations were as high as 14 mg/ml in fish in the sampling areas from the Balearen and Barbate. These were mostly in fish with a fork length greater than 18 cm. In comparison fish from Mazarron had plasma vitellogenin concentrations up to 4.6 mg/ml, however these were values for single specimens and the majority of values were below one mg/ml. For fish from the Bari sampling maximum values of 6.3 mg/ml were measured, although some of the Bari animals were larger than 18 cm fork length. In contrast to the plasma samples from a female fish samples from male fish had a much lower vitellogenin concentration as shown in Fig The highest values of.47 mg/ml were again shown in animals taken from Barbate or Baleares. In Mazarron the maximum concentration of vitellogenin was a.5 mg/ml. Correlation between Plasma Vtg and Season Length (cm) Although the samples were from different sampling areas and based on only a few days of the year one is still able to see an increase in the plasma vitellogenin concentration in female fish from May to July as shown in Fig This is similar to findings from previous studies. 127

128 Males,6,5 Mazzaron Bari Barbate Balearen Vtg Plasma (mg/ml),4,3,2,1, Length (cm) Figure 3.48 Relationship between plasma Vtg levels ( mg ml -1 ) and Fork length (cm) in male tuna of different sampling areas. 12 Females Vtg Plasma (mg/ml) Mazzaron Bari Barbate Balearen 2 Feb Mrz Apr Mai Jun Jul Aug Date 23 Figure 3.49 Seasonal changes in plasma Vtg levels in female fish in 23 Interestingly Fig. 3.5 shows the relationship between date of capture and plasma vitellogenin concentration for male fish. Higher values are correlated with May however this may be due to the size and fish and not due to seasonality. 128

129 ,6 Males Vtg Plasma (mg/ml),5,4,3,2 Mazzaron Bari Barbate Balearen,1, Feb Mrz Apr Mai Jun Jul Aug Figure 3.5 Seasonal changes in plasma Vtg levels in male fish in 23 Correlation between Plasma Vtg and GSI Date 23 Due to the fact that as a female fish matures GSI values increase and therefore there is a positive correlation with the vitellogenin concentration. As can be seen from Fig this relationship differs between sampling areas. Females Mazzaron Females Bari 5 5 Vtg Plasma (mg/ml) a) b) y=1,46 x,62 r ²=,86 Vtg Plasma (mg/ml) y=,33 x,6 r ²=, GSI GSI Females Barbate Females Balearen 4 12 Vtg Plasma (mg/ml) 3 2 c) d) Vtg Plasma (mg/ml) y=7,67 x+31,44 r ²=,5 1 y=11,92 x,59 r ²=, GSI GSI Figure 3.51 Relationship between plasma Vtg levels and GSI in female fish of four different sampling areas: a) Mazarron, b) Bari, c) Barbate and d) Balearen 129

130 Although there were variances in the significance of the positive correlation this was all too due to the maximum range of vitellogenin concentrations of plasma of the various sampling areas. Whereas in most cases the female fish from the different sampling areas showed positive correlations, although some were less significant due to the small sample size, the and amale fish did not show a positive correlation with GS I (Fig. 3.52). Which again indicates that body size is the most important factor controlling vitellogenin plasma concentration through accumulation. 1, Males Vtg Plasma (mg/ml),8,6,4 Mazzaron Bari Barbate Balearen,2, GSI Figure 3.52 Relationship between plasma Vtg and GSI for male fish from 23 fishing season Comparison of Vtg and ZRP in Wild Fish Using the new antibodies provided for ZrP it was now possible to make a comparison for sampling the wild populations over the three-year period. In general the values in male fish were extremely low if not below the detectable limit. Barbate From Fig 3.53 a good correlation can be found between Vtg levels and ZrP levels in the Barbate sampling. Again seasonality is an important factor with low levels in samples taken in April and May and an increase in June. However since Barbate traps catch fish which are on the way towards the Balearic Islands it is to be expected that the vitellogenin should be low. As previously stated all the values in the males studied were lower than.1 mg/ml and are shown only for control purposes here. 13

131 Barbate 3 Vtg ZrP 4. Male Female Vtg or ZrP [mg/ml] C13 C14 C12 C21 C23 C1 C6 C9 C3 C25 C17 C22 C29 Barbate 4 Vtg ZrP Male Female Vtg or ZrP [mg/ml] C4 C42 C38 C53 C49 C44 C54 C55 C41 C36 C37 C39 C35 C43 C5 C45 C46 C48 C51 Barbate Vtg ZrP 4. Male Female C61 C68 C64 C65 C62 C7 C66 C67 C69 C63 Vtg or ZrP [mg/ml] Figure 3.53 Plasma vitellogenin and ZRP values for the Barbate traps from The maximum values shown in this small sample of fish over the three years was 4 mg/ml indicating that these fish are on their way to the spawning ground and full vitellogenesis has not taken place. 131

132 Balearic Islands Vtg Female Balearen 4 Male 4, May June July Vtg [mg/ml] 3,5 3, 2,5 2, 1,5 1,,5, F-1 F-2 F-3 F-9 F-6 F-8 F-21 F-11 F-12 F-1 F-22 F-13 F-14 F-23 F-2 F-16 F-17 F-19 F-32 F-31 F-27 F-33 F-28 F-34 F-37 F-29 F-3 F-4 F-39 F-47 F-48 F-52 F-53 F-54 F-55 F-59 F-65 F-64 F-66 F-7 F-69 F-71 F-4 F-5 F-7 F-18 F-15 F-25 F-26 F-36 F-35 F-38 F-41 F-42 F-43 F-45 F-44 F-49 F-46 F-51 F-5 F-61 F-56 F-58 F-57 F-62 F-6 F-63 F-67 3,5 Female Balearen 5 May June July Male 3, 2,5 Vtg [mg/ml] 2, 1,5 1,,5, F F F F F F F F F - 27 F F F F F F F F F F F F F F F - 24 F - 21 F F F - 27 F F F F F F - 3 F F F F F F F F F F F - 16 F F F F F F F F - 17 F F F F F F F F F - 23 F F F F - 29 F - 29 F F F F F F - 46 F F - 47 F - 26 Figure 3.54 Plasma vitellogenin and females and male fish captured around the Balearic Islands in seasons 24 and 25. In contrast to some of the data from 23 the levels observed in 24 and 25 (Fig. 3.54) were markedly lower. This could be due to the sampling bias by the using long lines, purse seiners etc or the exact date of sampling. Maximum values around 3 mg /ml were observed in contrast to values above 178 mg/ml in 23. Again differences in sampling areas, sampling time and GSI of individual fish need to be taken into account if comparisons are to be made. The trends shown by vitellogenin were closely followed by those of ZRp as shown in Fig.3.55 and again males showed only background levels. 132

133 Balearen Female May June July Male 2.5 ZrP [mg/ml] F-1 F-2 F-3 F-9 F-6 F-8 F-21 F-11 F-12 F-1 F-22 F-13 F-14 F-23 F-2 F-16 F-17 F-19 F-32 F-31 F-27 F-33 F-28 F-34 F-37 F-29 F-3 F-4 F-39 F-47 F-48 F-52 F-53 F-54 F-55 F-59 F-65 F-64 F-66 F-7 F-69 F-71 F-4 F-5 F-7 F-18 F-15 F-25 F-26 F-36 F-35 F-38 F-41 F-42 F Female Balearen 5 May June July Male ZrP [mg/ml] F F F F F F F F F - 27 F F F F F F F F F F F F F F F - 24 F - 21 F F F - 27 F F F F F F - 3 F F F F F F F F F F F - 16 F F F F F F F F - 17 F F F F F F F F F - 23 F F F F - 29 F - 29 F F F F F F - 46 F F - 47 Figure 3.55 Plasma and ZrP concentrations in female and male fish from samples taken around the Balearic Islands in 24 and 25 during May and June and July. The central Mediterranean The sampling carried out in the central Mediterranean area was spread over four different areas in 23. In 24 only samples from Sardinia and the North Ionina Sea were available and in 25 no samples. Male fish again showed very low values, near to baseline and the maximum values in females between 3 and 4 mg/ml (Fig. 3.56). This is not perhaps surprising since all these sites lie outside spawning areas, although the higher values around the South Thyrrhenian sea are indicative of increased development and may perhaps be correlated with a spawning site. The close correlation between vitellogenin and ZrP could again be observed. 133

134 Vtg ZrP Golf of Taranto 23 Sardinia Channel 1,5 Female (May) Male Female (June) Male 1,,5, B 3 B 4 B 5 B 6 B 7 B 1 B 2 B 8 Vtg or ZrP [mg/ml] B 9 B 212 B 22 B 23 B 24 B 28 B 21 Vtg 5 4 ZrP South Thyrrhenian Sea (Sicily) 23 Strait of Messina Female (April) Female (June) Male Male Vtg or ZrP [mg/ml] Vtg or ZrP Concentration (mg/ml Females Sardinia Traps 25 Males North Ionian Sea F M. Vtg ZrP B42 B421 B428 B432 B433 B436 B437 B438 B44 B448 B422 B423 B424 B425 B426 B427 B429 B43 B431 B434 B435 B439 B441 B442 B443 B444 B445 B446 B447 B449 B45 B451 B412 B413 B414 B411 B415 Fish Nr. Figure 3.56 Plasma and ZrP concentrations in female and male fish from samples taken from the central Mediterranean in 24 and

135 Mazarron Cages For comparative purposes the data for the caged fish from are presented below. This includes measurements of both vitellogenin and ZrP in plasma. 3 Female Mazarron Vtg ZrP = control Vtg or ZrP [mg/ml] 2 1 zz2a zz31 zz33 zz35 zz36 zz38 zz39 zz11 zz4 zz41 zz43 zz51 zz52 zz53 zz55 zz56 zz11 zz15 zz18 zz113 zz114 zz115 zz116 zz118 zz119 zz121 zz122 zz123 zz125 zz24 zz26 zz28 zz211 zz212 zz216 zz217 zz218 zz22 zz223 zz224 zz225 zz226 zz229 zz231 zz234 zz235 Figure 3.56 Plasma vitellogenin and ZrP concentrations measured in captive female fish held in Mazarron from 23 until 25. In 24 and 25 fish were hormonally induced, control indicates fish which were not induced in this period. All these measurements were carried out in 25 therefore deterioration of the probes could have taken place. From the summary it is clear that very little gonadal development took place in 23 (Fig 3.56). In 24 both vitellogenin and ZRP were slightly higher, however in 25 maximum levels for ZRP and vitellogenin did not exceed 2 mg/ml. Normally during the spawning season levels should be much higher than this. In 23 the maximum GSI observed in the females sampled was around 3.. In 24 a similar value was reached but again with a vitellogenin/zrp value of around 1 mg/ml. This process repeated itself for 25 with similar values. However in general the GSI values were low with an overall average of The question now arises as to the cause of these low vitellogenin/zrp values. Either estradiol values are not reaching high enough levels to promote full vitellogenin production in the liver, or the genes responsible for vitellogenesis are not being switched on. The reasons for this may be many-fold but stress and nutrition within the cage structures may play an important role. In 25 postovulatory follicles were observed in many of the female fish this may indicate the spawning has occurred and vitellogenin levels have already decreased to the resting state. This is however unclear as the BFT is thought to be a multi-spawner. 135

136 3.6.4 Conclusions [11KT(ng/ml)] Ketotestosterone in wild Fish May June July Female Male [E2(ng/ml)] Estradiol in wild Fish May June July Female Male Figure 3.57 Comparison of 11-KTand Estradiol levels in both male and female wild fish in 25 for the months of May, June and July. For the steroid hormones there is a clear difference in the levels of 11 KT between male and female fish. A similar correlation is also shown for Estradiol with a complimentary peak concentration in June as would be expected (See Fig. 3.57). On making the comparison between the wild fish and cage fish as shown in figure 3.58 the general trends are reconfirmed. The estradiol levels in the wild fish were generally lower although the standard deviation was large. On looking back through the results for 23 and 24 it would appear that the cage fish do not suffer from too low estradiol levels, as the levels measured were comparable with the wild fish. Interestingly the values fur 11-KT and testosterone were higher in the cage fish, especially in terms of the males. This could be indicative of other changes within the reproductive hormone axis of the cage fish which requires further discrimination. A full analysis of all the samples for seasons 23, 24 and 25 might reveal a general trend in captive fish. 136

137 Estradiol: Mazarron 25 vs. Wild Fish 5 4 [E2(ng/ml)] Females Males Treated Control Wild Fish [T(ng/ml)] Testosterone: Mazarron 25 vs. Wild Fish Treated Control Wild Fish Female Male [11-KT(ng/ml)] Ketotestosterone: Mazarron 25 vs. Wild Fish Female Male Treated Control Wild Fish Figure 3.58 comparison of all three steroid hormones in male and female fish from treated control and wild fish populations in 25. The wild fish are average values taken from the Spanish sampling sites. 137

138 In the present study most of the GS I levels were below 3. excluding animals from the Balearic Islands, and known spawning area. The Barbate fish from the various traps possibly indicate fish on the way to the spawning grounds, and therefore the higher plasma vitellogenin concentrations. The Mazarron fish and the majority of the fish from Italian waters showed lower vitellogenin concentrations. One point of concern is the small size of the Mazarron fish. Although we know from previous studies that female tuna can reach sexual maturity at fork lengths greater than 135 centimetres, most of the Mazarron animals were below 18 centimetres. This could have adverse effects on their spawning ability. A correlation should be attempted to be made Vitellogenin concentrations and gonad staging. When different sampling areas are considered for the wild fish populations then a specific time frame should be applied to avoid seasonal variations. It is clear that samples from the Southern Tyrrhenian sea showed higher vitellogenin concentrations than those from the Straits of Messina. Again migration patterns, spawning areas and seasonal differences are to be expected. The correlation between tissue and plasma Vtg concentration is weak however for sex determination vitellogenin appears to be a good indicator as is the ZRP concentration. Females Mazarron 5 1,8 Mean Vtg & ZrP [mg/ml] 1,6 1,4 1,2 1,,8,6,4,2 N=6 N=2 N=4 N=5 Vtg ZrP, 1,-1,5 1,5-2, 2,-2,5 2,5-3,3 GSI Figure 3.59 Comparison vitellogenin and ZRP concentrations of plasma from Mazarron female fish together with their GSI values in 25. Although the total GSI levels in Mazarron fish were relatively low there was a good correlation between Vtg/ZRP levels and the GSI value itself. The absolute values however were extremely low throughout all the seasons measured (Fig. 3.59). In the first season in 23 this could be attributed to stress caused by multiple sampling throughout the year and the lack of feeding. In the second year 24 feeding was better and the stress levels experienced by the fish should have been significantly lower. In the 25 season all efforts were made to maintain the stress levels as low as possible and maintain feeding levels as high as possible. The gonadal development in the Mazarron cage animals was limited and this was reflected by the levels of the lipoproteins. From the analysis of the wild population, especially from the spawning areas some idea of the levels of vitellogenin/zrp experience in the wild was obtained. The monitoring of the cage population revealed low vitellogenin/ ZrP levels throughout the whole experimental period. These reinforce the general trend that the average GSI levels in the cage population were extremely low. They were however high enough to allow the fish to spawn. In captivity there appears to be a number of factors which influence vitellogenesis these need to be examined in more detail. 138

139 CHAPTER 4: SPAWNING ACTIVITY OF BFT IN THE MEDITERRANEAN SEA: WILD VS. CAPTIVE BROODSTOCKS 4.1 ABSTRACT The bluefin tuna (BFT, Thunnus thynnus) has become an endangered species due to overfishing. In order to resolve the contradiction between environmental and fishery needs, BFT was chosen as a prime candidate for domestication. Aiming at inducing successful breeding in captive BFT, the current study focused on determining the reproductive status of mature BFT captured at the Balearic Islands and held in floating cages on the coast of Murcia. BFT-specific immunoassays for measuring the profiles of central regulators of reproduction, i.e. gonadotropin-releasing hormones (GnRHs) and gonadotropins (GtHs) were developed and used to assess the reproductive potential of captive broodstocks. For this purpose, fish were sampled at three characteristic stages during the reproductive season, May, June, and July (23). Morphometric parameters (i.e. body weight, gonad weight) were recorded for individual fish, and blood and tissues (i.e. pituitary, brain, and gonads) were collected for further analyses. Our findings indicate that the major reproductive hormones, namely: sbgnrh, cgnrh-ii, and LH, all peak in June, during which the highest gonadosomatic index (GSI) values (av. 1.5±.3) were recorded both in captive and wild BFT. These results suggest that June is the natural spawning period of BFT, and that the reproductive endocrine system in captive-bft is functioning and preparing for the natural spawning season. This notion was further confirmed by comparing the aforementioned hormone levels to those of wild BFT collected during June along the Maltese coastline (a natural BFT spawning ground). No significant (P>.5) difference was observed in the levels of pituitary LH and native GnRHs in both captive and wild BFT populations. The fact that gonadotropins accumulate normally in the pituitaries of captive BFT emphasizes the reasoning for development of GnRH-based spawning induction treatments for this species. However, although gametogenesis did take place in captivity, the gonadosomatic index (GSI) values of the fish as well as the plasma levels of vitellogenin were relatively low compared to wild fish. This stresses out possible dysfunctions during the process of vitellogenesis resulting in the production of a smaller number of oocytes, and during spermatogenesis resulting in either reduced proliferation or increased apoptosis of spermatogonia. 4.2 DEVELOPMENT OF TOOLS FOR VERIFYING THE REPRODUCTIVE STATUS OF CAPTIVE BFT BREEDERS Introduction In recent years BFT farming was developed around the Mediterranean coast. However this kind of farming involves capturing of wild fish and their fattening in floating cages, therefore, does not conform to a typical aquaculture. Considering the growing demand for BFT and the increasing restrictions regarding its fishing, there is a great interest in developing specific procedures for controlling BFT reproduction in captivity. In view of the above, and given that development of methods to induce successful reproduction in captive fish has been feasible only through a basic understanding of the species' reproductive endocrinology (Zohar and Mylonas, 21), in the current study immunoassays for verifying the profiles of pivotal regulators of reproduction, i.e. gonadotropin-releasing hormones (GnRHs) and gonadotropins (GtHs), were established. These tools enabled assessing the reproductive statues of captive fish (sub-chapter 4.2), as well as evaluating the effectiveness of the GnRHa-based therapy (chapter 5). Three distinct forms of GnRH, namely: chicken GnRH-II (cgnrh-ii), salmon GnRH (sgnrh) and sea bream GnRH (sbgnrh), have been characterized in the brains of all perciform fish studied to date (Powell et al., 1994; Gothilf et al., 1995; White et al., 1995; Weber et al., 1997). Based on the resemblance of tuna reproductive features to that of several well-studied perciform fish, it is expected that the same identity of GnRH forms is conserved in BFT; however no information on this subject is available so far. Vertebrate reproduction is regulated by a coordinated action of the pituitary GtHs, with follicle-stimulating hormone (FSH) playing a dominant role in the initiation of gametogenesis and regulation of gonadal growth, and luteinizing hormone (LH) mainly dominating gonadal maturation and spermiation/ ovulation. Both, LH and FSH are heterodimers, sharing a common α-subunit and differing in their β-subunits. Nevertheless, gonadotropin duality in fish was controversial, and only during the last decade the existence of FSH was confirmed among gnathostomata species (Yaron et al., 23). Two distinct GtHs, FSH and LH, were isolated from the pituitaries of tuna species (Koide et al., 1993; Okada et al., 1994; Garcia-Hernandez et al., 1997), which share the highest sequence identity with the respective hormones of other perciform fish, i.e., 139

140 striped bass, tilapia and sea bream. In addition, the pituitaries of immature BFT were found to express both GtHs, each in a distinct gonadotropic cell type (Kagawa et al., 1998; Rodriguez-Gomez et al., 21), resembling the situation found in tilapia (Melamed et al., 1998) and sea bream (Elizur et al., 2). However, despite the isolation and characterization of tuna LH and FSH, until now no assays have been developed to monitor these hormones in tunas, and therefore, no information is available concerning their profiles during the reproductive cycle or in response to hormonal manipulation Materials and Methods RNA isolation, reverse transcription and amplification of sequences encoding for BFT native GnRH forms and GtH β-subunits Total RNA from brain or pituitary was isolated by the guanidinium thiocyanate phenol chloroform extraction method (Chomczynski and Sacchi, 1987) using TRIzole reagent (Gibco-BRL, Gaithersburg, USA). The extracted RNA (5 µg) was treated with 6 units of DNAse-RNAse free (Promega, Madison, WI, USA), and following DNAse inactivation (15 min at 65 C) the RNA was used to prepare the 5`-and 3`-RACE-ready cdnas by the SMART TM -RACE PCR (Clontech, Palo Alto, CA) according to the manufacturer's instructions. The obtained cdnas were subjected to RACE-PCR amplifications using primer sets consisting of the apposite anchor primer and gene specific primer (GSP; Table 4.1). For initial cloning degenerate GSP were designed according to amino acid sequences displaying high conservation among perciform species. Table 4.1 Gene specific primers (GSP) used to clone the cdna sequences encoding for BFT GtH β-subunits and native GnRHs. Primer Identification Primer Sequence (5` 3`) Primer Location (amino acid position) tufsh-f1 CAYGAYGARCARAARAT 45 5 tufsh-f2 TGYAAYGGNGAYTGG tufsh-r1 GGACGGACAGCTGGGTACG tufsh-f3 GAATTCGGGCAGGGTTGCAGTTAC FSHβ tufsh-r2 GCGGCCGCTTAATGATGATGATGATGATG AAAGGACGGACAGCTGGGTAC tulh-f1 GAATTCCAGCTGCCGCCCTGT 1 5 LHβ tulh-r1 GCGGCCGCTTAATGATGATGATGATGATG GTAGTAGAAAGGGATGTCATT CGnRH-II cii-7,2-f1 cii R1 CARCAYTGGTCNCAYGGNTGG CTGIGGYCTCARGTAGCTGC sgnrh s-5-2 -F1 s-3-2-r1 GTTGTTGGCGTTGGTGG CTCTCTTGGGTTTGGGC (-)11 (-) sbgnrh sbg-f1 sbg-r1 CGTCGACCAGCAYTGGTCITATGGIYTNAG GCTCCTCGACACAGCCCAGAACACTGC The identifications F and R denote primer direction: Forward (5 3`) and Reverse (3` 5`), respectively. Underlined letters represent the additional 6 histidine codones. Bold letters within primer sequences represent the following degeneracy: I- inosine; K- G or T; N- any of the four nucleotides (A/T/C/G); R- A or G; Y- C or T. Small uppercase letters indicate the EcoRI and NotI restriction sites. 14

141 The PCRs were carried out in a final volume of 5 µl using 2.5 units Taq polymerase (Promega), reaction buffer (Promega), 1.5 mm MgCl 2, nucleotides (.2 mm final concentration for each nucleotide), 25 pmol of each gene specific primer (listed in Table 4.1.1). Cycling parameters were: 3 min denaturation at 94 C followed by 3 cycles of 1 min denaturation at 94 C, 1 min annealing at 55 C and 1 min extension at 72 C. PCR products were purified with QIAquick PCR Purification Kit (QIAGENE, Hilden, Germany) cloned into pgem -T easy vector (Promega), and sequenced with ABI PRISM 31 Genetic Analyzer (Applied Biosystems, Foster City, USA) at the DNA Biological Services, Tel Aviv University, Israel. Gene identity was confirmed by comparing the obtained sequences with those available at the genebank ( Construction of Pichia pastoris expression vectors The cdnas encoding for the mature BFT LHβ and FSHβ were PCR amplified using the respective set of primers tulh-f1/tulh-r1 and tufsh-f3/tufsh-r2 (Table 4.1.1). Each of the anti-sense primers introduced an additional sequence codifying for six histidine residues (6xHis) flanked by a stop codon. The latter 6xHis tail tags the recombinant protein and facilitates its purification by an affinity column. Each PCR amplicon was introduced into the Pichia pastoris expression vector, ppic9k (Invitrogene, Carlsbad, CA), as an EcoRI / NotI insertion, 5` flunked by the sequence coding for the yeast mating factor-α secretion signal [S] Transformation and selection of BFT LHβ and FSHβ recombinant clones The constructed plasmid (5µg), encompassing the S-BFT-LHβ and S-BFT-FSHβ translational fusion were linearized with SalI and used to transform the host strain GS115 his4 (auxotrophic for histidine) by electroporation. The procedure was carried out by the Gene Pulser II Electroporation System (Bio-Rad, Hercules, CA) using the pulse parameters of 1.5 kv and 4 ohm, as established by transformation efficiency tests. Following selection on histidine-deficient agar plates, geneticin hyper-resistance transformants were picked for further expression analysis Production and purification of recombinant BFT LHβ and FSHβ Each selected colony was grown on buffered BMGY medium (1% yeast extract; 2% peptone; 1 mm potassium phosphate, ph 6.; 1.34% yeast nitrogen base; 4x1-5% biotin; 1% glycerol) in a shaking incubator (25 rpm) at 28ºC, for 2 days. The cells were harvested, re-suspended in buffered BMMY medium (same as BMGY but containing 1% methanol instead of 1% glycerol) to induce the AOX1 promoter, and grown for 6 days. The induced culture supernatant (about 2 liters) was concentrated through 1K MWCO membrane using the VIVACELL ultrafiltration system (Vivascience, Hannover, Germany), and washed several times with PBS ph 7.4, to eliminate the high phosphate concentration of the medium. Then the Histagged rbft-fshβ and rbft-lhβ proteins were purified by HiTrap chelating HP column (Amersham- Pharmacia-Biotech, Uppsala, Sweden) Antibody production Antibodies against recombinant rbft-fshβ were raised in two rabbits. The purified recombinant rbft-fshβ protein (5 µg) was dissolved in 1 ml of.9% NaCl and emulsified with complete Freund s adjuvant (1 ml; Sigma, Ness Ziona, Israel). Each rabbit received four subcutaneous injections at 3-week intervals. Test bleedings, to determine antiserum titers, were carried out 2 weeks after the third injection. The rabbits were bled 2 weeks after the final injection and the serum was aliquoted and stored at -7 C SDS-PAGE and Western blot analyses Pituitary protein samples were subjected to polyacrylamide gel electrophoresis using the SE 25 Mighty Small apparatus (Hoefer, San Francisco, CA). Then, proteins were electro-transferred from the SDS PAGE to a nitro-cellulose membrane using the TE 22 Mighty Small Transphor Unit (Hoefer). Nonspecific binding sites were blocked with 5% skimmed milk in PBS (137 mm NaCl, 8 mm Na2PO4, 1.4 mm KH2PO4, and 6.6 mm KCl, ph 7.5). For detection of FSH and LH, the membrane was incubated with the respective polyclonal antibodies either rose against rbft-fsh and stlh, diluted 1:5,, for 1 h at room temperature, washed five times with PBS-T (PBS plus.5% Tween-2), and then incubated with horse-raddish peroxidase (HRP) conjugated to goat ani-rabbit-igg. After 1 h of incubation, the membrane was washed as described above and the signal was visualized by an ECL detection system (Biological Industries, Beit Haemek, Israel) Statistical analysis 141

142 The results for each treatment are expressed as mean ± SEM. Statistical analyses were performed using Prism 4 (Graph-Pad Software, Inc., San Diego, CA). Homogeneity of variance was assessed by Bartlett s test and data were compared by ANOVA followed by Student-Newman-Keuls multiple range test Results Cloning of the cdnas encoding for BFT GtH β-subunits Superimposition of the 5` and 3` ends (Fig. 4.1A ) indicated that it comprises the full-length BFT FSHβ cdna (562 bp) including: 5` UTR (139 bp), putative signal peptide (45 bp; 15 aa), mature peptide (39 bp; 13 aa) and 3`UTR (69 bp) Interestingly, upon screening the clones encompassing 5` RACE amplicons, we identified a shorten version of the 5`UTR (18 bp vs. 139 bp). This would suggest the occurrence of an alternative splicing of the first intron, as was found in the tilapia FSHβ gene (Rosenfeld et al., 21) and in the chinook salmon LHβ gene (Xiong and Hew, 1991). Similarly, we cloned the full length cdna encoding for BFT LHβ cdna (639 bp), including: 5` UTR (48 bp), putative signal peptide (96 bp; 32 aa), mature peptide (348 bp; 116 aa) and the 3`UTR (111 bp) (Fig. 4.1B). Sequence comparisons of amino acids deduced from the respective BFT LHβ and FSHβ cdnas (Fig. 4.2) reviled high conservation of both molecules among tuna species (>96%). Yet, a wider comparison with cognate molecules derived from other Perciformes indicated conservation of the LHβ and diversification of the FSHβ Cloning of the cdnas encoding for BFT GnRHs Using BFT brain total RNA, GSP (listed in Table 4.1), and the RACE PCR cloning technique (described above), we have cloned the full length sequences of sgnrh and sbgnrh, and the partial sequence of cgnrh-ii (Fig. 4.3). These sequences reveal high homology rate (>91%) with sequences of closely related Perciformes including Dicentrarchus labrax (Gonzales-Martinez et al., 21), Sparus aurata (Gothilf et al., 1996), and Oreochromis niloticus (sbgnrh # AF

143 A ' TGC TAA CGA GGC AAA TAC AGA GCT CTA ATG GAA GCA AGC AGC AGA GTG ACG GTG M E A S S R V T V CAG GTG TTG TTG TTG GCG TTG GTG GTT CAG GTC ACC CTG TCC CAG CAC TGG TCC Q V L L L A L V V Q V T L S Q H W S TAT GGA TGG CTA CCA GGT GGA AAG AGA AGT GTG GGA GAG CTG GAG GCG ACC ATC Y G W L P G G K R S V G E L E A T I AGG ATG ATG GGC ACA GGG GTG TGG TAT CTC TTC CTG AAG AGG CGA GTG CCC AAA R M M G T G V W Y L F L K R R V P K CCC AAG AGA GAC TTA GAC CAT ACA ATG TAA TTA ATG ATG ATT CCA GTC ATT TCA P K R D L D H T M * L M M I P V I S ACC GAA AGA AAA GGT TCC CTC ATA AAT GAA GAG CTC C 3' T E R K G S L I N E E L C ' CAG CAT TGG TCA CAC GGC TGG TAT CCT GGA GGC AAG AGG GAG CTG GAC TCT TTT Q H W S H G W Y P G G K R E L D S F GGC ACG TCA GAG ATT TCA CAG GAG ATC AAG CTA TGT GAG GCA GGG GAA TGC AGC G T S E I S Q E I K L C E A G E C S B ' CTA ATA CGA CTC ACT ATA GGG CAA GCA GTG GTA TCA ACG CAG AGC ACG CGG GGA L I R L T I G Q A V V S T Q S T R G GTA TCG GCT TCA ACA AAG GAC AGC AGC CAA GAA CAA ATC AGA GAA GCA GCT TGC V S A S T K D S S Q E Q I R E A A C CTT ATG CAC AGA AGA ATG GCT ATG CAA ACC CTG GCA CTG TGG TTG CTG CTT CTG L M H R R M A M Q T L A L W L L L L GGC TCA GTG GTG CCA CAG GTC TGC TGT CAG CAC TGG TCA TAC GGA CTG AGC CCA G S V V P Q V C C Q H W S Y G L S P GGA GGG AAA AGG GAA CTG GAC AGC CTT TCA GAC ACA CTG GAC AAT GTA GTT GAG G G K R E L D S L S D T L D N V V E GGG TTT CCA CAT GTG GAC ACA CCT TGC AGT GTT TTG GGT TGT GTT GAG GAA TCG G F P H V D T P C S V L G C V E E S CCT TTT GCC AAA ATC TAC AGA ATG AAA GGA TTT CTT GGC AGT GTG ACC AAC AGG P F A K I Y R M K G F L G S V T N R GAG AAT GAA CAC AAA AAT TAT AAA AAA TGA TGA TTA TTT GAT TCT ACA ATA AAT E N E H K N Y K K * * L F D S T I N AAT CAT ATT GGC ATA TCT AAA AAA AAA AAA AAA AAA AAA 3' N H I G I S K K K K K K K TAC TTG AGA CCT CAG AGG AGG AGT CTT CTG AGA AAC ATT CTT TTG GAT GCC TTA Y L R P Q R R S L L R N I L L D A L GCC AGA GAG CTC CAG AAG AGA AAG TGA CAG CTT TCC ACC CTT CAC TGC TTT TCT A R E L Q K R K * Q L S T L H C F S ACT GAG TGA CCT GAC TCT CCT CTT 3' T E * P D S P L Fig 4.1 cdna sequences and deduced amino sequences of BFT FSHβ (A) and LHβ (B). The N-terminal of the mature peptide was designated position +1 and the amino acids in the signal peptide are given negative numbers. The nucleotide sequences of sense primers (F) and anti-sense primers (R) are marked with gray and black backgrounds, respectively. The polyadenylation signal is underlined. 143

144 FSHβ BFT GQGCSYGCHPKNISISVESCGITEFILTTICEGQCYLEDPVYISHDE***QKICNGDWSYEVKHIEGCPVGVTYPVARNCECTTCNTGNTYCGRLPGYVPSCPSF ThuOb 99% ---C---C C C---C ***---C C C-C-AC------C C--- Bonito 94% ---C---C----V--V---C------F---C---C ***---C C C-C-VC------C------T--CS-- Stb 8% ---C-F-C--T----Q---C-L--V-Y---C---C-H--L-----Y-RPE-R-C K-C C-C--C--E--DC--F-EDI--CL-- Sb 71% ---CRF-C--T-V-MP---C-G--V-Y---CA--C-H HDLAE-RTC D-C-LA CKC-MC-----DC-LFL-NI-TCLP- T 58% E-DC-S-CR-----LP-DTC-***-VD---C---CFQK--NF-HT-DWPK--TC--E------YTEQC-R-FI-----KC-C-AC-*A--DC-T-S--I--C*** LHβ BFT FQLPPCQLINQTVSVEKEGCASCHPVETTICSGHCITKDPVIKIPFSKVYQHVCTYRDFYYKTFELPDCPPGVDPTVTYPVALSCHCGRCAMDTSDCTFESLQPDFCTNDIPFYY ThOu 99% -----C C--C C---C C C C-C--C------C CM Bonito 98% -----C C-TC C---C C C C-C--C------C CM Stb 93% -----C CPKC C---C N-----C----LH C Q-C-C--C------C N-CM Sb 87% -----C CPKC C---C------M-TRY**-----C----LH C V-CSC-LC------C N-CM T 96% -----C CP-C C---C N-----C----L C C-C--C------C----M----CM Fig. 4.2 Sequence comparison of Perciformes GtH β-subunits. Sequences are aligned from the first amino acid of the mature peptide. Amino acids that were found to be identical to the related BFT sequence (in bold) are marked with dash, and the respective homology percentages are indicated on the left panel. Gaps (shown by an asterisk) were introduced to maximize alignment. The conserved 12 half-cysteine residues are marked with gray background. The abbreviation and sequence origin: Bonito- Koide et al., 1993; Sb- seabream, Elizur et al., 1996; Stb- striped bass, Hassin et al., 1995; T- tilapia, Rosenfeld et al., 1997; ThOu- Thunnus obesus, Okada et al.,

145 A ' TGC TAA CGA GGC AAA TAC AGA GCT CTA ATG GAA GCA AGC AGC AGA GTG ACG GTG M E A S S R V T V CAG GTG TTG TTG TTG GCG TTG GTG GTT CAG GTC ACC CTG TCC CAG CAC TGG TCC Q V L L L A L V V Q V T L S Q H W S TAT GGA TGG CTA CCA GGT GGA AAG AGA AGT GTG GGA GAG CTG GAG GCG ACC ATC Y G W L P G G K R S V G E L E A T I AGG ATG ATG GGC ACA GGG GTG TGG TAT CTC TTC CTG AAG AGG CGA GTG CCC AAA R M M G T G V W Y L F L K R R V P K CCC AAG AGA GAC TTA GAC CAT ACA ATG TAA TTA ATG ATG ATT CCA GTC ATT TCA P K R D L D H T M * L M M I P V I S ACC GAA AGA AAA GGT TCC CTC ATA AAT GAA GAG CTC C 3' T E R K G S L I N E E L C ' CAG CAT TGG TCA CAC GGC TGG TAT CCT GGA GGC AAG AGG GAG CTG GAC TCT TTT Q H W S H G W Y P G G K R E L D S F GGC ACG TCA GAG ATT TCA CAG GAG ATC AAG CTA TGT GAG GCA GGG GAA TGC AGC G T S E I S Q E I K L C E A G E C S TAC TTG AGA CCT CAG AGG AGG AGT CTT CTG AGA AAC ATT CTT TTG GAT GCC TTA Y L R P Q R R S L L R N I L L D A L B ' CTA ATA CGA CTC ACT ATA GGG CAA GCA GTG GTA TCA ACG CAG AGC ACG CGG GGA L I R L T I G Q A V V S T Q S T R G GTA TCG GCT TCA ACA AAG GAC AGC AGC CAA GAA CAA ATC AGA GAA GCA GCT TGC V S A S T K D S S Q E Q I R E A A C CTT ATG CAC AGA AGA ATG GCT ATG CAA ACC CTG GCA CTG TGG TTG CTG CTT CTG L M H R R M A M Q T L A L W L L L L GGC TCA GTG GTG CCA CAG GTC TGC TGT CAG CAC TGG TCA TAC GGA CTG AGC CCA G S V V P Q V C C Q H W S Y G L S P GGA GGG AAA AGG GAA CTG GAC AGC CTT TCA GAC ACA CTG GAC AAT GTA GTT GAG G G K R E L D S L S D T L D N V V E GGG TTT CCA CAT GTG GAC ACA CCT TGC AGT GTT TTG GGT TGT GTT GAG GAA TCG G F P H V D T P C S V L G C V E E S CCT TTT GCC AAA ATC TAC AGA ATG AAA GGA TTT CTT GGC AGT GTG ACC AAC AGG P F A K I Y R M K G F L G S V T N R GAG AAT GAA CAC AAA AAT TAT AAA AAA TGA TGA TTA TTT GAT TCT ACA ATA AAT E N E H K N Y K K * * L F D S T I N AAT CAT ATT GGC ATA TCT AAA AAA AAA AAA AAA AAA AAA 3' N H I G I S K K K K K K K GCC AGA GAG CTC CAG AAG AGA AAG TGA CAG CTT TCC ACC CTT CAC TGC TTT TCT A R E L Q K R K * Q L S T L H C F S ACT GAG TGA CCT GAC TCT CCT CTT 3' T E * P D S P L Fig 4.3 cdna sequences and deduced amino acid sequences of BFT sgnrh (A), sbgnrh (B), and cgnrh-ii (C). The GnRH decapeptide is framed. The nucleotide sequence of PCR primers is indicated with an arrow. 145

146 Optimization and validation of ELISAs to measure BFT native GnRHs In order to expedite the study on the reproduction in BFT we have previously demonstrated the ability of heterologous GnRH ELISAs to detect the respective hormones in tuna species (Rosenfeld et al., 23). Fig. 4.4 represents displacement curves for standard sbgnrh, cgnrh-ii and sgnrh that were used in this study to quantify the counterpart hormones in BFT pituitary extracts. LOGIT (B/B) Fig. 4.4 Displacement curves for standard sbgnrh, cgnrh-ii and sgnrh. The LOGIT function was utilized to transform standard curve to a linear plot. Each point is the mean of two determinations Optimization and validation of ELISA to measure BFT LH Considering the high conservation of the LHβ molecules among Perciformes, we could adopt with minor modifications an available ELISA, originally developed to measure striped bass LH (stblh; Mañanós et al., 1997), and validated to measure LH in several tuna species (Rosenfeld et al., 23). The introduced modification consisted of replacing native stblhβ with recombinant BFT LHβ (rbft LHβ). Fig. 4.5 represents displacement curves for standard rbft-lhβ and stblh. The parallel curves indicate immunological similarities between the native and the recombinant proteins. LOGIT (B/B) sbgnrh cgnrh II sgnrh 2 2 R 2 =.99 R 2 =.99 R 2 = y = Ln(x) R 2 = y = Ln(x) R 2 = Ln (Dose) Ln Dose -1 native stblhb rbft-lhb גול) stblhb. )native גול) )rbft-lhb. Fig. 4.5 Validation of the use of stblh ELISA for the rbft-lhβ. Displacement of standard stblhβ by rbft-lhβ in serial dilution Development of ELISA to measure BFT FSH Western blot analysis of BFT pituitary proteins indicated that the antisera generated in rabbits against rbft- FSHβ has high specificity to native BFT FSH (Fig. 4.6). Following hybridization with the anti-bft-fshβ (left panel), the analyzed membrane was striped and re-hybridized with anti stblhβ (right panel). Neither, native LHβ nor rbft-lhβ immuno-reacted with the anti-bft-fshβ. The anti-bft-fshβ specificity was further tested in an ELISA assay carried out according to Mañanós et al. (1997). The linear range of the calibration curve was maximized by optimizing the relative levels of coated standard rbft-fshβ and the primary antibody (Fig. 4.7). A coating concentration of 2 ng/ml, a primary antibody dilution of 1:5, (anti- BFT- FSHβ), and a secondary antibody dilution of 1:5 (goat-anti rabit IgG horseradish peroxidase conjugate) were determined to be the optimal conditions for this direct competitive ELISA. 146

147 Anti-stb LHβ Anti-BFT FSHβ kda Fig. 4.6 Detection of BFT GtH -subunits by anti-stblh (left) and anti-bft- FSHβ (right). BFT pituitary proteins (Lanes 1-4) and rbft- LHβ (Lane 5) were separated on SDS-PAGE (1%) and analyzed by Western blotting using anti-bft-fshβ. Then the membrane was striped and re-hybridized with anti-stblhβ. The corresponding positions of the molecular mass (kda) markers run simultaneously are indicated in between the immunoblots. OD Coating antigen 8 ng/ml EC ng/ml OD Coating antigen 4 ng/ml EC ng/ml log[concentration, M] log[concentration, M] OD Coating antigen 2 ng/ml EC ng/ml log[concentration, M] -.2 Fig. 4.7 ELISA calibration curves using various levels of rbft-fshβ as coating antigen. Coating with a 2ng/ml standard rbft-fshβ, followed by incubation with 1:5 primary antibody (anti-bft-fsh) revealed the the optimal ELISA conditions. Abbreviation: EC5; half maximum displacement Discussion In the current study we developed specific ELISAs for measuring native GnRH forms and GtHs in BFT. The attained ability to profile these hormones, which are all considered to be pivotal regulators of vertebrate reproduction, aids better understanding of BFT reproductive behavior. The cdnas encoding for three native GnRH forms were isolated from BFT brains. Sequence analyses reviled their identity as: sbgnrh (GnRH1), cgnrh-ii (GnRH2) and sgnrh (GnRH3), confirming the multiplicity and resemblance of the GnRH system operating in all perciform species studied to date (Powell 147

148 et al., 1994; Gothilf et al., 1995; White et al., 1995; Weber et al., 1997). The high conservation rate (>91%) of GnRH peptides across perciformes enabled the optimization and use of available generic ELISAs (Holland et al., 1998), to measure the respective hormones in BFT. Similarly, the cdnas encoding for the GtH β- subunits were isolated from BFT pituitary. These sequences were introduced into the methylotrophic yeast P. pastoris expression system for the production of the respective recombinant BFT-LHβ and BFT- FSHβ proteins. The employed system was proved to produce a satisfying yield of recombinant proteins in the correct immunological form. Consequently, these proteins were used as antigenic agents for the production of specific antibodies as well as standard proteins, both essential for the development of hormone specific ELISAs. The highly sensitive ELISA for measuring BFT LH was used already in the course of the REPRODOTT project and has helped: (i) studying the reproductive characteristic of captive BFT brooders and compare them with those of wild fish (sub-chapter 4.3), and (ii) evaluating the responsiveness of BFT endocrine system to the GnRHa implantation as well as the reliability of the tagging system to give notice of successful implantation (sub-chapter 5.2). The recent established ELISA for measuring BFT-FSH, albeit requires further optimization, is expected to extend our knowledge regarding the specific roles of FSH in BFT. It is important to note that thus far, ELISAs for measuring FSH are available only for salmonid species, all exhibiting synchronous ovarian development. Therefore, it is of great importance to have such a tool for characterizing FSH in BFT, daily spawner that exhibit asynchronous ovarian development. Beyond gaining valuable basic knowledge, our ability to profile the levels of FSH, a key hormone known to trigger and maintain gametogenesis in fish, should help identifying the concrete cause(s) for the deficient gametogenic development seen in captive BFT (subchapter 4.3), thus minimize the trial and error involved in establishing the breeding management for this species. 4.3 STUDYING THE REPRODUCTIVE BEHAVIOR OF WILD VS. CAPTIVE MEDITERRANEAN BFT Introduction Over the years BFT has attracted many research efforts that expanded the knowledge on the general biology, ecology, population dynamics, and fisheries management of this species (reviewed by Mather et al., 1995; Fromentin and Powers, 25). Yet, most of the data referring to BFT spawning season and spawning areas are largely based on two basic criteria: the presence of individuals with ready-to-spawn gonads, and/or the presence of eggs and larvae in the area. The earliest recognition of BFT larvae in the Mediterranean dates back to the investigations of Sella (1924, 1929) and Sanzo (1932) who found BFT larvae in Sicilian waters. More information regarding BFT spawning was obtained later on by using macroscopic scales of gonad maturation (Rodríguez-Roda, 1964a,b, 1967), biometric indices like the Index of Maturity (Sarà,1963, 1973) or Gonadosomatic Index (GSI; de la Serna and Alot 1992), and histological analysis of the gonads (Baglin, 1982; Corriero et al., 23; Karakulak et al., 24) as well as quantitative analysis of the reproductive steroid hormones and vitellogenin (i.e. a glycolipophosphoprotein precursor of the yolk proteins), whose concentrations in plasma or muscle vary according to the reproductive state of the fish (Heppell and Sullivan, 2; Bridges et al., 21; Susca et al., 21). The above-mentioned studies indicate that BFT reproduce during spring-summer. The western stock spawning season is from April-July in the Gulf of Mexico and along the Florida straits (Richards 1976; Montolio and Juarez 1977; Rivas 1978; Baglin 1982). The eastern stock reproduces in the Mediterranean during the months of May-July (Corriero et al., 23; Abascal et al., 24; Karakulak et al., 24; Fromentin and Powers, 25). In the current study, we further extended the knowledge regarding reproductive characteristics of both wild and cage reared BFT by verifying the profiles and dynamic interplay of key hormones along the brainpituitary-gonadal (BPG) axis that regulates reproduction. Since BFT was chosen as a prime candidate for domestication, and bearing in mind that when reared in captivity most fish exhibit some degree of reproductive dysfunction (Mylonas and Zohar, 21), such a comparison between wild and captive broodstocks should aid identifying potential blockage(s) along the BPG endocrine axis and facilitate breeding management protocols to overcome the recognized obstruction Materials and methods Experimental animals and husbandry The BFT used in the experiments were obtained from the spawning areas around the Balearic Islands, Spain, using a purse seine during the fishing campaigns of 22 and 23 (May June). Fish were maintained in a floating cage 5-m in diameter and 2-m in depth (adaptation cage) at the facility of Tuna Graso, S.A. in La Azohia, Spain, and fed to satiation once a day for 6 days a week with raw fish. Water 148

149 temperature inside the temperatures was recorded daily at 6 and 12 meters of depth from April 23 to July 25 (Fig. 1.5). The cage reared BFT were sacrificed in 23, 24 and 25 (Tables 4.2 and 4.3). During 23, a total of 24 fish were sacrificed (17 females and 7 males) in three characteristic stages of the reproductive cycle: May (4 females and 2 males), June (7 females and 3 males) and July (6 females and 2 males). In 24, 24 animals (7 females and 1 males) were sampled during late June. In 25, 14 fish (7 females and 7 males) were sacrificed during early July. Wild BFT were captured during 23 (May throughout July) at the Levantine Sea, Malta and Balearics. These fish were used to follow gonadal development in BFT aggregating around their Mediterranean spawning grounds (for more details see chapter 2). Additional wild fish were captured on June 2 th and 22 nd 24 by a commercial long-liner operating in the waters around Malta, for hormonal analyses (Table 4.4). From each fish fork length (FL) was measured to the nearest cm, total body weight (W B ) and gonad weight (W G ) were measured at the nearest kg and 1 g, respectively. Gonadosomatic index (GSI) was calculated as 1 W G W B -1. Table 4.2 Identification number, biometric data and histological characteristics of captive BFT ovaries. Identification number Month FL (cm) W B (kg) W G (g) GSI Most advanced oocyte stage* Postovulatory follicles May early vitellogenesis no 33 May late vitellogenesis no 35 May early vitellogenesis no 36 May early vitellogenesis no 11 June early vitellogenesis no 22 June perinucleolar no 27 June late vitellogenesis no 29 June late vitellogenesis* no 4 June late vitellogenesis* no 41 June early vitellogenesis no 43 June late vitellogenesis no 51 July lipid no 52 July perinucleolar no 53 July perinucleolar no 55 July perinucleolar no 56 July perinucleolar no 13 July perinucleolar no June late vitellogenesis no 18 June late vitellogenesis* no 114 June late vitellogenesis yes 115 June late vitellogenesis no 116 June late vitellogenesis no 118 June late vitellogenesis* no 123 June lipid no July late vitellogenesis* no 211 July late vitellogenesis* no 217 July late vitellogenesis yes 22 July late vitellogenesis* no 224 July early vitellogenesis no 225 July late vitellogenesis no 226 July late vitellogenesis* no FL, fork length; W B, total body weight; W G, gonad weight; GSI, gonadosomatic index; *, major α atresia of late vitellogenic oocytes (>5% atresia of late vitellogenic oocytes). 149

150 Table 4.3 Identification number, biometric data and histological classification of captive BFT testes. Identification Month FL (cm) W B W G (g) GSI Histological Spermiating** number (kg) classification* May early spermatogenesis no 34 May early spermatogenesis no 9 June 14 46, late spermatogenesis no 42 June late spermatogenesis no 44 June late spermatogenesis no 5 July spent no 54 July spent no 24 1 June late spermatogenesis no 12 June late spermatogenesis no 13 June early spermatogenesis no 16 June late spermatogenesis yes 17 June late spermatogenesis no 19 June early spermatogenesis no 11 June early spermatogenesis no 111 June late spermatogenesis yes 12 June late spermatogenesis no 124 June spent no July spent no 21 July late spermatogenesis no 213 July late spermatogenesis no 214 July early spermatogenesis no 219 July late spermatogenesis no 237 July late spermatogenesis no 238 July late spermatogenesis no FL, fork lenght; W B, total body weight; W G, gonad weight; GSI, gonadosomatic index; *The histological classification was based on the types of spermatocysts observed in the germinal epithelium and the abundance of spermatozoa in the lumen of seminipherous tubules (further details are reported in the text). **Fish were classified as spermiating if they released semen either spontaneously when on board, or in response to a pressure on the abdomen Hormonal and histological analyses Captive and wild BFT were sampled as described above for hormonal and histological analyses. Each fish was lifted onto the deck of a service boat using a hydraulic crane, and blood was collected either from the lateral artery or directly from the heart. Pituitaries were removed, immediately frozen in liquid nitrogen, and stored at 8ºC until further analysis. Pituitary extracts were prepared according to Holland et al. (1998), with proportionally doubling the volume of the extraction buffers. The LH levels were measured in pituitary extract and plasma using the adapted ELISA for BFT LH (sub-chapter 4.2.2). Levels of all three GnRHs (sgnrh, sbgnrh, cgnrh-ii) were also measured in BFT pituitary extracts using the respective highly sensitive ELISAs (Rosenfeld et al., 23). For the evaluation of reproductive status, ovarian and testicular fragments from a total of 55 captive-reared and 8 wild BFT specimens were analyzed histologically. Gonads slices were fixed in Bouin s fixative or 1% buffered formalin, dehydrated in increasing ethanol concentration, clarified in xylene and embedded in paraffin wax. Five-µm thick sections were cut and stained with Haematoxylin-Eosin. For the classification of reproductive state of females, the most advanced oocyte stage was recorded for each specimen, according to the classification of Corriero et al. (23). The presence of atretic and postovulatory follicles was also recorded. Vitellogenic atretic follicles were classified at α or β stage according to Hunter et al. (1986). For the classification of reproductive state of males, the type of spermatogenic cysts was recorded and the quantity of spermatozoa in the lumen of seminipherous tubules was subjectively evaluated Results Seasonal profiles of GnRHs and GtHs in wild and captive BFT During 23, captive BFT were sampled at three consecutive months (May, June and July), spanning the reproductive season of the Eater Atlantic BFT stock. Our results indicate that the pituitary LH content was 15

151 greatest (P<.5) in females collected in June (Fig. 4.8A). A similar LH peak was also observed in males; however, it was not significant due to non-representative sample. Interestingly, our measurements indicate exceedingly higher levels (2-3 orders of magnitude; mg vs. ng per gland) of LH within BFT pituitaries as compared to the recorded quantities for smaller Perciformes (i.e., seabream: Holland et al., 1998; striped bass: Hassin et al., 1998). Yet, the LH levels in BFT plasma (Fig. 4.8B) were found to be on the same scale (ng/ml) with other Perciformes. Pituitary LH (mg/gland) A A B A, B F M May June July Plasma LH (ng/ml) B F M May June July Fig. 4.8 Pituitary LH content (A) and plasma LH levels (B) in cage-reared BFT (23). Levels (mean ±SEM) are expressed as total amount (mg) per pituitary and ng per ml plasma, respectively. Groups with different superscripts are significantly different (P<.5). In all samples, the pituitary levels of sbgnrh (Fig. 4.9A) were 1-fold higher as compared to cgnrh-ii (Fig. 4.9B). The levels of sgnrh were below the detection limit of our assay (<.5 pg/well) and are therefore not shown. Similarly to the pituitary LH profile, the levels of both sbgnrh and cgnrh-ii significantly peaked during June. 151

152 sbgnrh (ng/pit) A. b a, b a May June July F M cgnrh-ii (pg/pit) B. a b a F M May June July Fig. 4.9 Pituitary content of sbgnrh (A) and cgnrh-ii (B) in cage-reared BFT (23). In accordance with the results obtained in 23, during the consecutive year (24), cage reared BFT as well as wild BFT caught along the south coast of Malta (a natural BFT spawning ground), were sampled during the third week of June. Our analyses indicate that the content of pituitary LH and native GnRHs (cgnrh-ii and sbgnrh) in cage reared BFT do not vary significantly (P>.5) from those found in wild population (Fig. 4.1). pituitary LH (mg/pit) A Control Wild pituitary sbgnrh (ng/pit) B Captive Wild Fig. 4.1 Pituitary LH content (A), pituitary cgnrh-ii levels (B) in captive and wild BFT populations (24). Hormone levels (mean±sem) are expressed as total amount per pituitary. Groups with different superscripts are significantly different (P<.5) Temporal pattern of gonad maturation The average GSI values of wild BFT caught during 23 around their natural Mediterranean spawning grounds peaked during June (Fig. 4.11A). Likewise, the GSI values of captive BFT peaked during June; 152

153 however, these maximal GSI values were almost 3-fold lower compared to those measured in wild fish (av ±.28 and 4.13 ±.11, respectively). Data collected during 24, further attest the significant (P <.1) enhanced gonadal development in wild compared to captive BFT (av ±.37 and 2.56 ±.29, respectively; Fig. 4.11B). Regression analysis using data from both genders demonstrated several positive linear correlations between GSI values and pituitary content of LH, sbgnrh, and cgnrh-ii (Table 4.4). The LH levels showed a higher correlation to GSI values (P=.1) and to sbgnrh levels (P=.26) than to the levels of cgnrh-ii (P=.5). A B GSI GSI values values (%) captive wild B *** ab *** a A *** b B GSI values (%) a b May June July Captive Wild Fig Gonadal development of wild and captive BFT sampled during 23 (A) and 24 (B). Different letters above bars indicate significant (P<.5) difference between means. Asterisks denote values significantly different between groups (*** P <.1). Table 4.4 Correlation, and corresponding coefficient (r 2 ), between GSI and pituitary content of LH and native GnRHs (sbgnrh and cgnrh-ii) n=21 LH(mg/pit) ccgnrh-ii (pg/pit) sbgnrh (ng/pit) GSI r 2 =.684 P=.1 sbgnrh r 2 =.485 P=.26 cgnrh-ii r 2 =.431 P=.5 r 2 =.497 r 2 =.17 P=.22 P=.46 r 2 =.32 - P= The histological evaluation of ovarian maturational stage in captive BFT is sumerized in Table 4.1. Among fish sampled in May 23, three individuals had oocytes at early vitellogenesis (Fig. 4.12a) and one showed also oocytes at late vitellogenesis (Fig. 4.12b). In June 23, one specimen had only perinucleolar stage oocytes (Fig. 4.12c), two had oocytes at early vitellogenesis and four showed also oocytes at late vitellogenesis. Among the latter fish, 2 displayed major α atresia (i.e. > 5% atresia of late vitellogenic oocytes) (Fig. 4.12d). All the fish sampled in July 23 had regressed ovaries, being the most advanced oocytes at the perinucleolar and lipid stage in 5 and 1 fish, respectively. In 24, one individual had the most advanced oocyte population at the lipid stage and six individuals had oocytes at the late vitellogenesis stage. Among the latter fish, 2 displayed major α atresia and one showed postovulatory follicles (Fig, 4.12e, f)). In 25, one out of seven fish had oocytes at early vitellogenesis as the most advanced development stage. All the other specimens had late vitellogenic oocytes and, among them, one had also postovulatory follicles and 4 showed major α atresia. 153

154 Fig Micrographs of the ovaries from captive BFT reared in floating cages at the coast of La Azohía (Murcia, Spain). (a) Ovary from an individual with early vitellogenesis as the most advanced oocyte stage. (b) Ovary from a specimen with oocytes at late vitellogenesis stage. (c) Ovary showing only perinucleolar stage oocytes. (d) Major α atresia of vitellogenic follicles. (e) Ovary with oocytes at late vitellogenesis stage and postovulatory follicles. (f) Higher magnification of part of (f) showing two postovulatory follicles. Haematoxylineosin staining. Bar = 3 µm in (a), (b), (c), (d), (e) and 15 µm in (f). Arrow, postovulatory follicle; af, α atresia of vitellogenic follicles; ev, oocyte at early vitellogenesis stage; lv, oocyte at late vitellogensesis stage. The histological evaluations of the testes of captive BFT are summarized in Table 4.2. In May 23, the two sampled males were at early spermatogenesis stage as their testes showed most of the germinal epithelium occupied by cysts containing spermatocytes and spermatids, but cysts containing spermatozoa and luminal spermatozoa were also observed in most of seminiferous tubules (Fig. 4.13a). All fish sacrificed in June 23 (n=3) had testes at late spermatogenesis as their germinal epithelium consisted mainly of cysts containing spermatids and spermatozoa, and the lumen of seminipherous tubules was filled with spermatozoa (Fig. 4.13b). The testes of both of the 2 fish analysed in July 23 were spent since the germinal epithelium consisted mainly of spermatogonia, and only few spermatocysts were visible, mainly with spermatids or spermatozoa; luminal residual spermatozoa could still be observed (Fig. 4.13c). Among the 1 males analyzed in 24, 3 individuals were classified at early spermatogenesis and six individuals had testes that were classified at late spermatogenesis. Two fish of this last group were in spermiating conditions, i.e. sperm was released from the testes when the fish were brought on board, either spontaneously or in response to external pressure. One specimens was classifies as spent. In 25, 1 individual was classified at early spermatogenesis, 5 at late spermatogenesis and 1 was spent. 154

155 Fig Micrographs of the testes of captive BFT. The fish were sacrificed in 24 and 25 during the natural spawning season of the species in the Mediterranean. (a) Testis from an individual classified at early spermatogenic stage showing germinal epithelium occupied by cysts containing spermatocytes and spermatids; luminal spermatozoa can also be observed. (b) Testis from a specimen classified at late spermatogenesis. The germinal epithelium consists mainly of cysts containing spermatids and spermatozoa and the lumen of seminipherous tubules is filled with spermatozoa. (d) Testis from a bluefin tuna classified as spent. The wall of seminipherous tubules is often devoid of spermatocysts and residual spermatozoa can be observed in the lumen of seminipherous tubules. Haematoxylin-eosin staining. Bar = 3 µm. arrow, germinal epithelium; l, lumen of seminipherous tubules; sz, spermatozoa. Histological evaluation of maturation stage of wild fish is summarized in Table 4.3. All females (n=6) captured around Malta in June 24 had late vitellogenesis as the most advanced oocyte stage and did not show signs of recent imminent spawning. One of the two wild male fish had quiescent testes characterized by a prevalence of intertubular tissue and a germinal epithelium constituted mainly by spermatogonia; the other individual had testes that were classified at late spermatogenesis Discussion Hormonal analyses of pituitaries collected from caged-reared BFT indicate that the major regulators of reproduction, namely: sbgnrh, cgnrh-ii, and LH, all peak in June, coinciding with highest GSI values. Likewise, studies that followed gonadal development in eastern BFT population favor June-July as the natural 155

156 spawning season in the Mediterranean Sea (chapter 2; Medina, 22; Abascal, 23; Corriero, 23). Taken together, these findings suggest the onset and function of BFT endocrine system under captivity. Moreover, a comparison of the reproductive characteristics of wild and captive BFT broodstock indicates that LH, the maturational-gonadotropin, accumulates normally in the pituitary of captive BFT. The fact that pituitary LH content is not a limiting factor, supports the employment of a GnRHa-based therapy for inducing breeding among captive BFT (chapter 5), as was successfully done with various cultured fish (reviewed by Mylonas and Zohar, 21; Zohar and Mylonas, 21). Nevertheless, at three consecutive years (23-25) considerable lower GSI values were recorded in captive BFT, compared to those found in wild populations. This notion was further attested by histological analyses showing slowed down maturational processes concomitant with high rates of apoptotic activities both in the ovaries and tests of captive fish. This points to possible dysfunctions during the process of vitellogenesis resulting in the production of a smaller number of oocytes, and during spermatogenesis resulting in either reduced proliferation or increased apoptosis of spermatogonia {Corriero, submitted #366}. Alternatively, an inappropriate broodstock diet might have resulted in a lower availability of the necessary nutrients for gametogenesis, resulting in reduced fecundity and sperm production, as demonstrated in other fishes. These findings underline the need to study gametogenesis and broodstock diets in the nearest future. 156

157 CHAPTER 5: MATURATION IN CAPTIVITY AND SPAWNING INDUCTION USING HORMONAL THERAPIES 5.1 INTRODUCTION Tuna farming is a capture-based aquaculture industry (Ottolenghi et al., 24), and involves the capture of wild fish and their fattening in floating cages for periods ranging between 2 months to 2 years. Fueled by the increasing demand for this unique fish by the sashimi-sushi market in Japan, Europe and the United States (Catarci, 24), the expansion of the fattening industry is considered threatening to the wild stocks, which is now considered to be overfished (Fromentin and Powers, 25). Therefore, there is a great interest in developing broodstock management, spawning and larval rearing methods, in order to support the sustainable development of bluefin tuna farming. Furthermore, studying the reproductive biology of this species in captivity would result in a better understanding of its life history traits, required for a better management of the wild stocks. Efforts at spawning the bluefin tuna were initiated in Japan two decades ago, and have been met with some success (Lioka et al., 2; Masuma et al., 24). Fish were maintained in large cages or enclosures and were allowed to spawn naturally. Spawning occurred without consistency, and for a decade after 1983, there was little progress as tuna stopped laying eggs, resulting in a limited success in establishing a reliable supply of good quality eggs in order to develop larval rearing methods. Eggs were once again obtained during the 199 s and in June of 22 an artificially reared adult bluefin tuna produced 1 million eggs for the first time at Kinki University (Japan Times, 24). This was a significant step forward and paved the way for the future of full-scale farming of this species. In fact, three 1 m-long tunas grown from these eggs were harvested at 2 kg and shipped to the market in September 24 (Japan Times, 24). Still, the failure of bluefin tuna in Japan to reproduce reliably over the years, underlines the existence of reproductive dysfunctions in fish reared in captivity. Similar to bluefin tuna, reproductive dysfunctions are exhibited by most fishes reared in captivity (Zohar and Mylonas, 21). In females, there is commonly a failure to undergo final oocyte maturation (FOM), ovulation and spawning; while a less serious problem in males is the reduction in semen quantity or quality. These dysfunctions probably result from the combination of captivity-induced stress (Sumpter et al., 1994; Pankhurst and Van Der Kraak, 1997), the lack of the appropriate "natural" spawning environment (Zohar, 1989a; Zohar, 1989b; Battaglene and Selosse, 1996) or even the lack of essential dietary components in the broodstock diet (Watanabe and Vassallo-Agius, 23). Reproductive dysfunctions often weaken as consecutive generations of broodstock are being produced from cultured parents, as fish are inadvertently selected for characteristics adaptive to the culture environment. For the development of an aquaculture industry for the bluefin tuna, as for any other species established so far, it is imperative to study its reproductive biology and establish methods for the induction of maturation and spawning in captivity (Bromage and Roberts, 1995). These methods may be useful not only during the early establishment of the industry, but they can prove to be important management tools in order to ensure efficient spawning and a consistent supply of gametes for the future. Part of the REPRODOTT study was an effort to develop methods for the artificial Control of reproductive maturation in captive-reared bluefin tuna. Specifically, the objectives were to (a) examine the maturation of wild-caught Mediterranean bluefin tuna under culture conditions, (b) produce and characterize a GnRHadelivery system, and develop an underwater administration method, (c) induce FOM, ovulation/spermiation and spawning and (d) study the effect of the GnRHa treatment on the reproductive endocrine axis, gonadal histology and gamete characteristics. 5.2 PREPARATION AND ADMINISTRATION OF GONADOTROPIN-RELEASING HORMONE AGONIST (GNRHA)-LOADED IMPLANTS AND INDUCTION OF OVULATION Introduction Manipulation of reproductive processes can be achieved using a variety of hormonal treatments (Zohar and Mylonas, 21). Of the available hormones, agonists of gonadotropin-releasing hormone (GnRHa) offer a great advantage due to their high potency, lack of species specificity and stimulation of endogenous gonadotropin release (Crim and Bettles, 1997; Peter and Yu, 1997). Treatment with GnRHa can be done in the form of simple injections or sustained-release delivery systems (Donaldson, 1996; Crim and Bettles, 1997; Zohar and Mylonas, 21). Due to the short half-life of GnRHa, a simple injection induces a brief elevation in plasma luteinizing hormone (LH), the gonadotropin responsible for FOM (Nagahama et al., 1994). As a result, some fishes require multiple GnRHa injections for an effective treatment, whereas in 157

158 some occasions the required handling for multiple injections is prohibitive, both for management and fish welfare reasons. This is especially true of wild broodstocks (Hodson and Sullivan, 1993) and fish with large body size, such as the greater amberjack (Seriola dumerili) (Micale et al., 1997; Lazzari et al., 2), dusky grouper (Epinephelus marginatus) (Marino et al., 23) and bluefin tuna (Mylonas, 23). Sustained-release GnRHa-delivery systems, on the other hand, induce long-term elevation of plasma LH with only a single treatment (Crim et al., 1988; Mylonas et al., 1998a; Mylonas et al., 1998b), and have proven effective in inducing FOM, ovulation and spawning, as well as enhancing spermiation in many fishes (Mylonas and Zohar, 21b). In the present study, the objectives were to (a) produce and characterize a GnRHa-delivery system for use with very large fish (>1 Kg), (b) develop an underwater hormone-implant administration method, and (c) induce FOM, ovulation/spermiation and spawning Materials and Methods Experimental animals and husbandry The bluefin tuna used in the experiments were obtained from the spawning areas around the Balearic Islands, Spain, using a purse seine during the fishing campaigns of 22 and 23 (May June). The required number of fish was first transfered into a floating cage 5-m in diameter and 2-m in depth (adaptation cage) at the facility of Tuna Graso, S.A. in La Azohia, Spain. For the experiments during the 24 spawning season, fish from the adaptation cage were allocated into two 25-m diameter cages (No 1 and 2) and one 5-m diameter (No 3), all 2 m in depth (Table 5.1). Cage No 1 contained fish from the 22 fishing campaign, placed in the cage in February 23. Fish in this cage were not handled or managed during the 24 experiments and were observed daily for spawning activity by placing an observer near the cage, from dusk to midnight during the natural reproductive season (June July). Cage No 2 contained fish from the 23 fishing campaign, placed in the cage in September 23. These fish were used to study the potential of GnRHa treatment in inducing spawning. Therefore, after GnRHa implantation (see later), fish were not handled any further and the cage was observed daily for spawning activity. Cage No 3 contained fish from the 23 fishing campaign, placed in the cage in February 24. These fish were used to study the effect of GnRHa implantation on FOM, ovulation/spermiation and spawning, and the associated changes in the reproductive endocrine axis. Therefore, a few days after treatment all fish in the cage were sacrificed in order to collect blood and tissues, and verify maturation (see later). For the 25 reproductive season experiments, the fish used were those from Cages No 1 and 2 of the 24 experiments described above (Table 5.1) and all fish were sacrificed a few days after GnRHa implantation. 158

159 Table 5.1 Design of the experiments on the induction of final maturation, ovulation/spermiation and spawning of captivereared bluefin tuna using GnRHa-implants. Year of experiment Cage No 1 (25-m diameter) Years in captivity Number of untreated fish (Control) 28 Males all 2 Females all Number of GnRHa-implanted fish Males none 6 Females none 4 Sex ratio (male/female) unknown 2. Implantation date n/a 3 June 25 Sample time (d after treatment) n/a 8 Cage No 2 (25-m diameter) Years in captivity Number of untreated fish (Control) 16 Males unknown 5 Females unknown 7 Number of GnRHa-implanted fish 12 Males unknown 1 Females unknown 6 Sex ratio (male/female) Implantation date 22 June 24 4 July 25 Sample time (d after treatment) n/a 2-3 Cage No 3 (5-m diameter) Years in captivity 1+ Number of untreated fish (Control) Males 1 Females 7 Number of GnRHa-implanted fish Males 3 Females 6 Sex ratio (male/female) 1. Implantation date 23 June 24 Sample time (d after treatment) n/a = not applicable Throughout the maintenance of the fish in the cages, they were fed to satiation once a day for 6 days a week with raw fish, which included Pacific mackerel (Scomber japonicus), Atlantic mackerel (Scomber scombrus), sardine (Sardina pilchardus), herring (Clupea arengus) and squid. Water temperature inside the temperatures was recorded daily at 6 and 12 meters of depth from April 23 to July 25 (Fig. 5.1). Mortalities were also recorded (5 fish during the three years of the study), and any dead fish were removed by divers. 159

160 6 meters 2 meters Mar Jun-23 Sep-23 Dec-23 Mar-24 Jun-24 Sep-24 Dec-24 Mar-25 Jun-25 Time (23-25) Fig. 5.1 Water temperatures at a depth of 6 m (gray line) and 2 m (black line) in the sea cages at La Azohia, Murcia, Spain, where the bluefin tuna broodstocks were maintained during the three years of the study. The black bars at the bottom indicate the natural reproductive period of bluefin tuna in the Mediterranean Sea, and the vertical arrows point to the time the ovulation induction experiments were carried out in 24 and Development of a GnRHa Enzyme-linked Immuno-sorbent Assay (ELISA) Plasma GnRHa levels were determined using a heterologous, competitive ELISA based on the protocol described for striped bass luteinizing hormone (LH) ELISA (Mañanos et al., 1997). The specific antibodies (AbGnRHa) used in the assay were raised in rabbits by a commercial company (Agrisera, Sweden), against the analogue desgly 1, dala 6, Pro 9 -GnRH-NEthylamide (Bachem, Switzerland), a commonly used GnRHa in fish. The ELISA included the following steps: 1) Coating the wells of 96-well microtiter plates (Nunc Maxisorp) with.1 ml well -1 of GnRHa solution (1 ng ml -1, diluted in sodium carbonate buffer.5 M, ph 9.6), overnight at 4ºC. The non-specific binding (NSB) was determined in wells (triplicate) coated with sodium carbonate buffer. After coating, the wells were washed (3x 1 min) with PBST (sodium phosphate buffer.1 M, ph 7.4, containing.9% NaCl and.5% Tween-2). Washing was repeated after every step of the assay. 2) Blocking the wells with.2 ml well -1 of PBST containing 2% of non-fat dry milk for 3 min at 37ºC, to reduce background. 3) Incubation of standard and samples with the specific antibodies (AbGnRHa at 1:8,), inside the coated wells, overnight at 4ºC. Before distributed into the wells (.1 ml well -1, in duplicate), the reaction mixture was prepared in test-tubes, in PBST-2%, and pre-incubated overnight at 4ºC. 4) Incubation with.1 ml well -1 of horseradish peroxidase (HRP)-labelled goat-anti-rabbit IgG (Biorad), diluted in PBST-2% at 1:1,, for 45 min at 37ºC. 5) Development of color with.1 ml well -1 of TMB reagent (Biorad) in complete darkness at room temperature for 25 min and stopping the reaction by addition of.1 ml well -1 of 1N sulphuric acid. Absorbances were measured after 5 min at 45 nm, using a microplate reader (Thermomax, Molecular devices, Menlo Park, CA). The expression of results was performed after linearization of the sigmoid standard curve, using the logit transformation (logit (Bi/Bo)=ln(Bi-NSB/Bo-NSB)), were Bi represents the binding of each point, Bo the maximum binding and NSB the non-specific binding. A series of tests was performed to optimize the assay protocol, studying the behavior of the standard curve under different temperatures and incubation times. Before analysis, plasma samples were extracted using a modification of the method described by Harmin and Crim (1993). For extraction, plasma was mixed (1:4) with methanol and vortexed vigorously. After centrifugation for 15 min at 16, g (4ºC), the supernatant was dried under a flow of gaseous Nitrogen and the moist pellet was re-suspended with distilled water. The extract was lyophilized, re-suspended in ELISA buffer (PBST), centrifuged for 15 min at 3, g (4 C) and the supernatant was stored at -2ºC until analysis with the GnRHa ELISA. 16

161 Validation of the ELISA for measuring GnRHa in fish plasma was done by demonstrating parallelism between serial dilutions of plasma from individuals treated with GnRHa and the standard curve prepared in ELISA buffer. The accuracy of the ELISA was tested by determining the recovery percentage from spiked samples. The reproducibility of the ELISA was tested by calculating the intra- and inter-assay coefficient of variation (CV). Intra-assay variability was assessed by conducting replicate measurement in a single ELISA (same microtiter plate) of samples (n=4) with a known quantity of GnRHa, chosen to be around the 5% binding. The inter-assay variability was calculated by analyzing 12 GnRHa doses, in 1 different ELISA undertaken at different days Preparation and validation of GnRHa-delivery systems The GnRHa-delivery systems (i.e., implants) were prepared by loading the agonist desgly 1, dala 6, Pro 9 - GnRH-NEthylamide (Bachem, Switzerland) into a matrix of poly [Ethylene-Vinyl Acetate] (EVAc) according to the procedure of Freese (1989), with some modifications. Briefly, 56 or 84 mg of GnRHa (24 and 25 experiments, respectively) and.16 g of bovine serum albumin (BSA, Sigma, Germany) were dissolved in 7.5 ml dh 2 O, and were mixed with 7.5 ml dh 2 O containing.16 g of Inulin (Sigma, Germany). The mixture was frozen at 8 C and lyophilized for 48h (Alpha 1-2, Martin Christ, Germany). The dried powder was ground using a glass tissue-grinding rod connected to an overhead mixer (RZR 22, Heidolph, Germany). Twenty ml of a 15% EVAc solution in MeCl 2 were then added to the GnRHa/BSA/Inulin powder, and the mixture was vortexed for 5 min and sonicated for 3s at 3 watts (UP 2S, dr. Hielscher GbmH, Germany). The produced emulsion was poured into a levelled aluminium cast (5 x 5 x 4 mm) placed on a block of dry ice. The solidified plate was then placed in a 2 C freezer for 3 days in order to evaporate the MeCl 2, followed by 48 h in a vacuum desiccator to remove any moisture. The implants were punched from the dried GnRHa/BSA/Inulin/EVAC plate using a 3 mm dermal punch (Keyes Punch 3mm, Miltex GmbH, Germany). The in vitro release kinetics of the GnRHa implants were evaluated using the procedure of Mylonas et al. (1995) and Sarter et al. (26). Briefly, the implants (n=4) were embedded in a 2% solution of low melting agarose (Sigma, Germany) in in vitro buffer (3.36 g KH 2 PO 4, g NaHPO 4,.4 g sodium azide,.4 ml Tween 8 in 2 l of dd H 2 O, adjusted to ph 7.), at the bottom of 7 ml flat-bottom p[ethylene] vials. Once the agarose solidified, 5 ml of the in vitro buffer was added and the vials were placed on a rocking shaker in an incubator at 24 C. At various times afterwards, a 5 µl sample of the in vitro buffer was taken from each replicate and stored at 8 C until analysis for GnRHa. The in vitro buffer was replaced with fresh one after each sampling. The release of GnRHa from the implants was also examined in vivo using 2-year-old shi drum (Umbrina cirrosa) as model fish (mean ± SD weight of 629 ± 111 g). Fish were maintained in two 2-m 3 tanks, supplied with water from a recirculation system (24 ± 1 C). Fish were fed ad libitum (Trouvit, Hendrix, Italy) using a demand feeder, and were starved for two days before sampling to reduce handling stress. Individuals from one group (n = 6) received a GnRHa implant, administered into the dorsal musculature using an implanting device, and the rest (Control) were simply punched with the implanting device. At various times after implantation, blood samples were collected from the caudal vasculature into heparinized syringes, centrifuged for 1 min at 13g (Microfuge, Heraeus), and the separated plasma was stored at 8 C until analysis for GnRHa GnRHa implantation For application to bluefin tuna in 24, two implants of 2.35 mg of GnRHa were used for each fish, to produce effective doses of about 8 4 µg GnRH Kg -1 body weight, since fish were expected to weight between 6 and 12 Kg. For the 25 experiments the dose was increased, and each fish was implanted with two implants of 3.96 mg of GnRHa, to produce effective doses of about 1 5 µg GnRH Kg -1, since fish were expected to weight between 8 and 16 Kg. The two implants were attached to a polyethylene arrowhead (Floy Tag and Manufacturer Company, USA) using a.5 mm nylon monofilament (Fig. 5.2A), which was passed through the center of the implants using a 2G needle. On the same monofilament, a 15- mm piece of fluorescent yellow tubing was added (reporter-spacer), which was used as an indicator of successful implantation. Another piece of coloured tubing was added (1 mm) on the nylon monofilament, which functioned as a quick identifier tag of a fish already implanted (see later). Finally, a three-position, three-color code was created using colored tubing, and was used to identify individual fish (27 individual combinations) after implantation. Based on the experience of the 24 experiments, the GnRHa implant assembly was modified slightly in 25. The monofilament was now first passed through a metal cone guide before going through the implants, in order to prevent them from rubbing against the skin and muscle block during insertion (Fig. 5.2B). In addition, a metal disk was placed after the implants in order to prevent them from being pushed over the yellow tubing during insertion. 161

162 A im sp tag a C a tag spear cod B im d c D Fig. 5.2 Preparation of GnRHa implants and implantation of captive-reared bluefin tuna. (A) The two GnRHa implants (im) were attached to a polyethylene arrowhead (a) via a.5 mm nylon monofilament, followed by a 1 mm yellow reporter-spacer (sp), a 1-15 cm pink tag (tag), and a three color individual identification code (code). (B) In the second year experiment (25), a modification was done on the implant assembly by the addition of a metal, conical leader (c) and disk stopper (d) around the implants (im), in order to prevent them from rubbing against the skin and muscle block of the fish during insertion, and being pushed over the yellow tubing. (C) The arrowhead was fitted on a specially prepared spear and was administered to the fish using a spear gun. (D) The implant was administered at the posterior part of the tail (arrows) and the success of the implantation procedure was first evaluated by the diver, after visually examining if the yellow reporter-spacer was inside the fish. The GnRHa-implantation experiments in both years were planned for the second half of the natural spawning season of bluefin tuna in the Mediterranean (Medina et al., 22; Abascal et al., 23; Corriero et al., 23), but for the 25 experiments an added prerequisite was the establishment of water temperatures above 23 C, the spawning temperature of bluefin tuna (Schaefer, 21). For the 24 experiments, an accessory cage (25 x 12 x 9m) was attached to Cage No 3 (5-m diameter) and 16 fish were herded inside by divers. The next day, a skin diver administered the GnRHa implants to all fish in the accessory cage, using a spear gun fitted with a specially designed spearhead (Fig. 5.2C). The GnRHa-implants were loaded onto the spear gun and the fish were implanted one-by-one. Compressed air (i.e., SCUBA) was not used, as it was found that the produced air bubbles were distressing the fish. The skin diver was instructed to target the lateral muscle area above the lateral line and posterior to the first dorsal fin (Fig. 5.2D), in order to prevent accidental penetration of the abdominal wall and damage to the internal organs. For each fish, the skin diver had to remain under water for a considerable amount of time, in order to approach the fish at the appropriate distance to enable correct placement of the GnRHa implant. After each implantation, the diver checked visually the implantation site on the fish and evaluated if the yellow reporter-spacer (Fig. 5.2A) was 162

163 inside the muscle of the fish. Appearance of the yellow reporter-spacer outside the fish would indicate that the implant did not penetrate deep enough into the muscle. Once the implantation was completed, fish were herded out of the accessory cage and returned to their cage until they were sampled/sacrificed after 5-6 days. The following day, the same procedure was repeated in Cage No 2 (25-m diameter), but without the use of the accessory cage, since this was a smaller cage and the fish could be crowded easier. Instead, the bottom of the cage was lifted in order to split the cage in half and to reduce the volume of the cage. Once the implantation was completed, the bottom of the cage was allowed to drop and the cage regained its full volume. No further manipulation was done to these fish and the existence of any breeding behavior and the presence of spawned eggs was monitored daily between sunset and midnight (see later). In 25, GnRHa-implantation was done using a similar procedure, but this time the bottom of the two cages (No 1 and 2) was raised higher, reducing the maximum depth to about 5 m, allowing the skin diver to get closer to the fish without having to dive too deep, as well as to target the fish from above. Based on the successful implantation experience from the previous year, in 25 the diver was instructed to again target the dorsal body mass, this time including the area around the dorsal fin. After GnRHa-implantation the existence of any breeding behavior and the presence of spawned eggs was monitored daily between sunset and midnight in Cage No 1, until the fish were sampled/sacrificed after 2-3 (Cage No 2) or 8 (Cage No 1) days GnRHa-implantation evaluation and gonad sampling Evaluation of GnRHa-implantation was done a few days after implantation by sacrificing all fish (Table 5.1). At this time, gonads were collected from both GnRHa implanted fish and un-implanted, Control fish in the same cage. Sampling was done by sacrificing one fish at a time using an underwater shot-gun. For sacrificing, the cages were handled in a similar way as for GnRHa implantation. Each fish was lifted onto the deck of a service boat using a hydraulic crane, and blood was collected either from the lateral artery or directly from the heart. Fork length, wet body weight and the weight of the gonads was recorded, and the first spiniform ray of the first dorsal fin was removed to determine the age of the fish (Cort, 1991). The piece of muscle surrounding the implant was photographed and then dissected out in order to evaluate the success of the implantation technique, by looking at the appearance of the yellow reporter-spacer and examining the presence of the GnRHa implants inside the muscle. The gonads were dissected from the fish onboard, placed on ice in a styrofoam box and transported to the laboratory within 1 h. For histological evaluations of maturation stage, ovarian and testicular fragments were fixed in a 4% formaldehyde, 1% gluteraldehyde buffered saline (McDowell and Trump, 1976) and embedded in glycol methacrylate resin (Technovit 71; Heraeus, Kulzer, Germany). Serial sections were obtained at a thickness of 3 µm on a microtome (Biocut 235, Reichert Jung, Germany) and were stained with methylene blue/ azure II/ basic fuchsin (Bennett et al., 1976). Spermiation and ovulation was verified onboard either by the presence of semen or eggs coming out of the genital pore after application of pressure on the abdominal area, or by the expulsion of semen or eggs after gentle pressure of the dissected gonads. Ovulated eggs were removed from the dissected ovaries by stripping into a clean, dry 4-l plastic bucket. Sperm collected earlier from a single male and stored on ice, was placed on top of the eggs and mixed together using a plastic spoon. At the same time, clean seawater was added, the eggs were swirled gently for 2 min, and were then allowed to stand for 5 min. After rinsing with plenty of seawater, the eggs were placed in a 2-l sealed container filled with 5 l of seawater, placed on ice in a styrofoam box and transferred to the laboratory facilities where the eggs were placed in conical incubators with water at a temperature of C Egg collection from cages In 24, a 1.5-m deep plastic curtain was placed along the circumference of Cage No 2, in order to prevent any spawned eggs from escaping, thus potentially allowing their collection using dip-nets or Bongo-net trawls over the surface of the cages, which was between sunset and midnight (Sawada et al., 25). Similar egg surveys were also done from the waters around Cage No 1. Due to failure of collecting any eggs in 24, the curtain was modified in 25, by fitting it with five tubes constructed of 5 µm mesh and having rigid PVC cylinders at their end, also fitted with a 5-µm mesh on their bottom. These tubular meshes were extending radialy from the curtain, and hence from the cage, following the direction of the water current. Eggs/larvae from the cage could be carried by the current into these tubular nets and trapped in the PVC cylinders. The PVCs cylinders were examined for eggs, and were cleaned every morning after sunrise and 163

164 every night before sunset. Eggs collected were taken to the laboratory incubated at 21 C in microtiter plates (Panini et al., 21), and their development was followed until yolk absorption Statistical analysis Data from individual fish on fork length, wet body weight, GnRHa levels, oocyte diameters, and sperm density and motility were analyzed by Analysis of Variance (ANOVA) at minimum significance of P <.5. Statistical analyses were done using a linear statistics software (Sigma Stat and SuperAnova, USA). When necessary to normalize distribution (sperm motility %) values were Arcsin-transformed prior to the ANOVA. Results are presented as means ± SEM, unless indicated otherwise Results Development of the GnRHa ELISA The optimal concentrations for the antigen and antibody were 1 ng GnRHa ml -1 for coating and 1:8, AbGnRHa dilution, respectively. The chosen assay conditions gave NSB less than 1% and maximum binding (Bo) around 1., which was well within the range of the lineal response of the microtiter plate reader (data not shown). The standard curve ranged from.2 to 12 ng ml -1, at binding between 8% and 2%, respectively (Fig. 5.3). The sensitivity of the assay was determined as 51.3 pg ml -1 by means of low detection limit, and calculated from Bo 2SD. Depending on the GnRHa concentration, intra- and inter-assay CVs varied between 2.% and 4.9%, and 1.8% and 9.8%, respectively, while spike recovery was 79 ± 9%. The suitability of the GnRHa ELISA for measuring GnRHa in plasma was demonstrated by the parallelism between the standard curve prepared in ELISA buffer, and serial dilutions of bluefin tuna and shi drum plasma implanted with GnRHa (Fig. 5.3) Sample dilution Standard curve bluefin tuna Shi drum GnRHa concentration (ng ml -1 ) Fig. 5.3 Parallelism between the GnRHa standard curve (Logit Binding %) prepared in ELISA buffer and binding curves obtained by serial dilutions of plasma extract from GnRHa-implanted bluefin tuna or shi drum In vitro and in vivo release of GnRHa from the developed implants The GnRHa release in vitro was maximal the first 2 d, and declined continuously thereafter (Fig. 5.4A). The mean initial release was 525 ± 166 µg GnRHa implant -1 day -1, while after 6 d the amount released was reduced to between 19 ± 52 and 17 ± 18 µg GnRHa implant -1 day -1. The plasma GnRHa profile in vivo reflected the release in vitro, as maximal plasma levels were again observed 1 d after implantation and statistically significant (two-way ANOVA, DNMR, P <.1) plasma levels were measured until day 7 (Fig. 5.4B). Thereafter, plasma GnRHa levels were not significantly different from un-treated Controls, though GnRHa was detectable by the ELISA. 164

165 J JJ A Fig. 5.4 (A) Mean (± SEM) daily release of GnRHa from the implants in an in vitro assay at 24 C, and (B) mean (± SEM) plasma GnRHa levels in an in vivo release study at 24 C using shi drum as a model fish. Asterisks indicate GnRHa means which were significantly different from Controls (ANOVA, DNMR, P <.1) GnRHa implantation J J J J J J J J Time (days) J J J GGG J G G G G In 24, the implantation process proved to be difficult and required an increasing amount of time as the day progressed, starting with 2 min and finishing at 3 min per implantation. This was due to the fact that the fish became aware of the activity of the diver rapidly, and developed an avoidance behavior. As a result, the diver had to approach the fish very slowly and wait until they calmed down and reduced their swimming speed around the cage, in order to move within firing distance. Furthermore, the fish responded rapidly to the sound made by the speargun being fired, by making quick horizontal turns, thus moving away from the trajectory of the spear. As a result, the incidence angle of the arrow was more than 9º resulting, in some cases, of failure of the arrowhead to penetrate adequately into the fish s muscle, resulting in the GnRHa implants either being outside the fish or just barely underneath the skin (Figs. 5.5A and B). The placement of the reporter-spacer functioned very well in assisting the diver to determine if the implantation was successful or not. Only in one fish the evaluation of the diver and the evaluation undertaken after sacrificing the fish differed, giving a reliability of 93%. On the other hand, verified also by the examination of the implants after sacrificing the fish, the implantation procedure was determined to be 64% successful. The actual numbers of GnRHa-treated males and females in Cage No 3 were 3 and 6, respectively, while 12 fish of unknown sex were implanted in Cage No 2 (Table 5.1). J ** J ** ** J J G GnRHa Control B GnRHa (ng ml -1 )

166 A reporter spacer tag implants B cod reporter C D E Fig. 5.5 Photographs of GnRHa-implant assemblies at the time of sacrificing the fish, showing the tag (tag) and three-color code (code) for identifying individual fish. (A) An implant in 24 that failed to penetrate well into the muscle block, evident from the fact that the yellow reporter-spacer was visible outside the body. (B) On evaluation in the laboratory after extraction of the muscle block, it was confirmed that the GnRHa implants remained outside the muscle. (C-E) Successfully administered GnRHa implants in 25, evident from the fact that the yellow reporter-spacer was not visible outside the body. In 25, the implantation procedure was much easier and progressed faster requiring an average (± SD) of only 3.1 ± 1.4 min for each fish, implanting a total of 16 males and 1 females in the course of 2 days (Table 5.1). As this year the bottom of the cage was lifted higher and the fish were targeted from above, the diver was closer to the fish and at a better position for targeting. In addition, due to the position of the diver, sharp avoidance turns made by the fish on hearing the speargun firing did not affect the angle of attack of the arrow, which was always at a right angle to the swimming direction of the fish. As a result, the majority of implants penetrated well into the muscle block (Figs. 5.5C-E). As in the previous year, the reliability of the diver evaluation regarding implantation success was very high (94%), whereas the implantation success was improved greatly, reaching 84%. The fork length and wet body weight of the fish treated with GnRHa implants was not significantly different (three-way ANOVA, P =.83) from the un-treated Controls (Fig. 5.6), indicating that the random implantation procedure did not result in a bias toward either smaller or larger individuals. As expected, the size of the fish in 25 was significantly greater (three-way ANOVA, P <.1) than in 24 (Fig. 5.6). In both years, males were significantly larger (three-way ANOVA, P <.1) than females, having a mean (± SD) of 8 ± 1% larger fork length, and 28 ± 1% larger body weight. The age of the fish used in the experiments ranged between 5 and 12 years, with males being significantly older than the females (three-way ANOVA, P <.1). The mean (±SEM) age in 24 and 25 of males was 7.5 ±.4 and 8.8 ±.3, and of females was 6.5 ±.3 and 7.6 ±.3, respectively. 166

167 225 2 A ANOVA Parameter P value Year.1 Sex.7 GnRHa implant.49 Control GnRHa Fork length (cm) B ANOVA Parameter P value Year.2 Sex.3 GnRHa implant Females 24 Males Females Year of experiment 25 Fig. 5.6 Mean (± SEM) fork length (A) and wet body weight (B) of captive-reared bluefin tuna males and females in 24 and 25 (n = 5-16, as shown inside the mean bars at the bottom graph), at the completion of the maturation induction experiments. Fish were maintained in floating cages for 1 year (24 experiment) or 2-3 years (25 experiments) prior to treatment with GnRHa implants. There were statistically significant differences (3-way ANOVA, P values on graphs) between years and sexes, with both fork length and wet body weight being greater in 25 compared to 24, and males being larger than females in both years. There were no statistical differences between Controls and GnRHa implanted fish Effect of GnRHa on plasma GnRHa levels and ovulation/spawning Males In 24, GnRHa could not be detected in the plasma of the GnRHa-implanted bluefin tuna, but in 25 there was a statistically significant (two-way ANOVA, P <.5) difference between GnRHa-implanted and Control fish (Fig. 5.7). The effect was similar between the two sexes (two-way ANOVA, P =.46). Wet body weight (Kg)

168 1.8.6 ANOVA Parameter P value GnRHa Implant.2 Sex Control GnRHa Females Sex Males Fig. 5.7 Mean (± SEM) plasma GnRHa of captive-reared bluefin tuna males and females in 24 and 25 (n = 7-16, as shown above each mean bar), 2-8 days after treatment with GnRHa implants. In both sexes, there was a statistically significant elevation (2-way ANOVA, P values on graph) in GnRHa implanted fish. Two Control males in 24 and five GnRHa-implanted males in 25 were found to be spermiating when they were sacrificed and brought on-board. Two females implanted with GnRHa 3 days before in 25 were found to have ovulated eggs in their ovaries, and the eggs were inseminated with sperm obtained from two of the spermiating males, before transport to the laboratory for further incubation. Eggs of bluefin tuna 19 h after in vitro fertilization had a mean (± SD) diameter of 143 ± 18 µm (Fig. 5.8A). Hatching begun 28 h after artificial fertilization (21-23 C) and the newly hatched pre-larvae measured 2.6 mm in total length, growing rapidly to 3.2 mm 8 h after hatching (Fig. 5.8B) and 3.6 mm 19 h after hatching (Fig. 5.8C). No spawning activity was observed during the monitored periods in either years, but eggs were collected from Cage No 1 in 25 (Fig. 5.8D), which produced larvae (Figs. 5.8E to G) identified as bluefin tuna (Alberto Garcia and Francisco Alemany, Istituto Español de Oceanografia, personal communication). 168

169 A D B E F C G Fig. 5.8 Macro-photographs of bluefin tuna embryos 19 h after artificial insemination (A), a pre-larva 36 h after fertilization or 8 h after hatching (B), and a pre-larva 19 h after hatching (C). Putative bluefin tuna embryos collected from Cage No 1, four days after GnRHa implantation (D), and the pre-larvae produced from these eggs after hatching in the same afternoon (E), 2 d after hatching (F) and 3 d after hatching (G). All photographs were taken using a 4x objective and the bar at the bottom of each photo represents 1 mm Discussion This is the first report of a hormonal treatment of bluefin tuna or any other thunnid species, resulting in the induction of FOM, spermiation and spawning, and underlines a real possibility for controlling reproduction in captive bluefin tuna, as well as other tunas of interest. The results could spark worldwide efforts at inducing spawning in wild-caught, captive-reared bluefin tuna which are widely maintained in many parts of the world by the fattening industry. In the course of the present study, an ELISA has been developed for the first time for an agonist of GnRH, and it has contributed significantly to the development and optimization of the GnRHa implants, in order to ensure that they release an adequate amount of GnRHa, and with the desired kinetics. The GnRHa ELISA can be of further use in future studies of controlling the reproductive function of bluefin tuna, in evaluating the effect of administration of the GnRHa implants in different broodstocks, and in ensuring that the released amounts of GnRHa in the fish s circulation are adequate to induce FOM, spermiation and spawning. Furthermore, the GnRHa ELISA can also be used in studies with other species, as it was demonstrated that it can measure GnRHa extracted from the plasma of at least two other species, the European sea bass (Dicentrarchus labrax) and the shi drum. More recently, the GnRHa ELISA developed here has also been 169

170 used to measure the release of GnRHa from sustained-release delivery systems in the Senegal sole (Solea senegalensis) (E. Mañanos, unpublished data). The produced GnRHa implants released high amounts of GnRHa for a period of 7 d, with smaller amounts being released for up to 28 d in the in vitro study. Such GnRHa-delivery EVAc implants were first developed for use in Atlantic salmon (Salmo salar) (Zohar, 1996) and later for gilthead seabream (Sparus aurata) and striped bass (Morone saxatilis) (Mylonas and Zohar, 21b). Due to the very large size of the bluefin tuna, it was necessary to modify the manufacturing of the implants, in order to allow for the extremely high amounts of GnRHa required by such fish. The release of GnRHa from the EVAc implants is a diffusion-controlled process, and it is the result of the slow dissolution of the GnRHa-bulking compound mixture, which is trapped in channels through the polymer matrix during the manufacturing process. The chemical composition of the bulking agent (a mixture of a water-soluble BSA and water-insoluble Inulin) determines the rate of dissolution of the bulking agent, as the amount of GnRHa included in the mixture is usually very small (1-13%). The release of GnRHa from implants used for smaller broodfish usually lasts between days (Mylonas and Zohar, 21b). However, in the case of the bluefin tuna, which have a body size between 1 to 4x larger than broodfish of other species, the amount of GnRHa included in the implant constituted 6% of its total bulking compound mixture. Due to the very high water solubility of the GnRHa, the dissolution of the GnRHbulking compound mixture upon administration into the fish was very high, thus the release rate of GnRHa was much faster than reported in other studies (Mylonas et al., 1997a; Mylonas et al., 1998a; Mylonas et al., 1998b). The release kinetics were also affected by the higher water temperature during the experiments with bluefin tuna (23 C) compared to other fish where similar GnRHa implants have been used so far (Sorbera et al., 1996; Zohar, 1996; Larsson et al., 1997; Mylonas et al., 1998b), further increasing the rate of GnRHa release by dissolution and diffusion. Still, a GnRHa release for more than 7 d from the developed implants was adequate to induce at least one cycle of FOM, ovulation and spawning 3 d after treatment, demonstrating that captive-reared bluefin tuna do respond to GnRHa therapy during the reproductive season. The greatest potential of sustained-release GnRHa implants in the induction of FOM, ovulation and spawning in multiple spawning females with a daily or almost daily ovulation/spawning frequency, such as the bluefin tuna. For example, females of the family Sparidae, such as the red porgy (Pagrus pagrus), red seabream (Pagrus aurata) and gilthead seabream (family Sparidae) have an asynchronous mode of ovarian development and are capable of spawning on a 24-h cycle for periods up to 4 months (Watanabe and Kiron, 1995; Zohar et al., 1995; Mylonas et al., 24b). In the gilthead seabream, while a single GnRHa injection at the onset of the spawning season induced daily spawning in only 2% of the fish, more than 7% of the females given a GnRHa implant continued spawning on a daily basis (Zohar et al., 1995). Similar results have been obtained with the red sea bream and red porgy (Matsuyama et al., 1995; Zohar and Mylonas, 21). In the present study, only a single spawning event has been documented by the collection of fertilized eggs, but the presence of post-ovulatory follicles together oocytes undergoing FOM demonstrates that another ovulation and spawning would occur shortly. Due to the fact that the fish were sacrificed, it was not possible to document the ovulation and spawning frequency of the fish, but it seems certain that more that one cycle of FOM, ovulation and spawning can be induced as a result of the GnRHa treatment, as is the case in many other asynchronous fishes. Underwater administration of hormone implants to non-anaesthetized fish of any species has not been reported to date. In the absence of appropriate, non-lethal handling procedures for large pelagic fishes such the bluefin tuna, the development in the present study of the hormone implantation method is an indispensable tool in the efforts of controlling reproductive processes in large pelagic fishes, and in developing hormone-based spawning induction procedures. The success of the implantation procedure increased significantly over the two years of the study, though it never reached 1%. Reducing the volume allowed for the fish to be swimming during the implantation, by lifting the bottom of the sea cages higher and reducing the depth of the cage during the second year of the study, increased implantation success from 63 in 24 to 83% in 25. With a shallower cage, the fish could be targeted from the surface of the water, which has two major advantages. Firstly, the skin diver was not limited in his targeting of the fish, by how long he could remain underwater, and was thus able to select a fish, follow its movement, target and then implant at the most opportune moment. Secondly, implanting the fish from the top meant that the fish had less optical contact with the diver and even if they did turn suddenly on hearing the firing of the speargun, this did not move them away from the projectile of the speargun and did not change the angle of attack of the spear, ensuring that the arrowhead hit them with enough force to allow penetration of the GnRHa implants. Based on the experience obtained in the present study, therefore, it is recommended that the fish are crowded in an area of 15-2 m 2 and 4-5 m in depth for implantation. If the fish are smaller that the ones used in the present study, then the crowded area can be reduced even more. 17

171 The vast majority of failed implantations during both years had to do with failure of the arrowhead to penetrate enough into the fish s muscle block, thus leaving the GnRHa implants outside the body. As mentioned above, this problem was solved to a great extend by reducing the depth of the cage and allowing the skin diver to target the fish (a) from above and (b) from a closer distance. Nevertheless, we believe that manufacturing a smaller arrowhead of a lesser diameter could improve implantation success, as the smaller arrowhead will require less force to break through the tough skin of bluefin tuna and be inserted into the muscle block. In addition, the damage done to the fish will be less, allowing for repeated treatments, either during the same reproductive season or in subsequent years. One of the major objectives of REPRODOTT was to determine if mature, wild-caught bluefin tuna would undergo reproductive maturation and spontaneous spawning if maintained in captivity. Reproduction of other thunnid species in captivity has been reported recently for the yellowfin tuna (Thunnus albacares) (Wexler et al., 23) and the Pacific bluefin tuna (Thunnus orientalis) (Sawada et al., 25). In the study of yellowfin tuna, wild-caught, mature fish were placed in outdoor tanks and begun spawning spontaneously after a few months, demonstrating that captivity did not inhibit reproductive maturation in this species. In the case of Pacific bluefin tuna, a closer relative to the Atlantic bluefin tuna of the present study, the captured individuals were immature (< 1 Kg) and underwent reproductive maturation and spontaneous spawning after rearing for 7 years in floating cages. However, spawning was not obtained every year and in one occasion 1 years passed before eggs were produced again from the captive broodstock (Sawada et al., 25). Based on the recorded stage of gonadal maturation of both male and female bluefin tuna during the natural reproductive season in the present study (see Chapter 5.4), it seems possible that these fish could undergo FOM, spermiation and spontaneous spawning in captivity. During the two years of the study, for example, Control males contained spermatozoa in their testes, although only in 24 two fish were producing significant amounts of sperm that could be obtained by stripping. Furthermore, almost all Control females were in full vitellogenesis but only two showed signs of ovulation. However, one could never know if fish at such stage of reproductive maturation would indeed complete their reproductive cycle and undergo maturation and spawning, since failure of females to undergo FOM and production of low amount of expressible sperm are the most common reproductive dysfunctions observed in captive fishes (Zohar and Mylonas, 21). Some examples in females include various flatfishes (Berlinsky et al., 1996; Larsson et al., 1997; Mugnier et al., 2), members of the Serranidae family (Tucker, 1994; Watanabe et al., 1998), the striped bass (Morone saxatilis) and white bass (M. chrysops) (Mylonas and Zohar, 21a), the fugu (Takifugu spp.) (Yang and Chen, 24; Chen, 25), the shi drum (Barbaro et al., 22; Mylonas et al., 24a) and the dusky grouper (Marino et al., 23). Some examples of males include the yellowtail flounder (Pleuronectes ferrugineus) (Clearwater and Crim, 1998), turbot (Scophtalmus maximus) (Suquet et al., 1992), certain strains of Atlantic salmon (Salmo salar) (Zohar, 1996) and Atlantic halibut (Hippoglossus hippoglossus) (Vermeirssen et al., 2). Therefore, maintaining mature, wild-caught bluefin tuna in captivity and simply waiting for them to undergo spontaneous FOM, spermiation and spawning would not be the fastest approach for the acquisition of fertilized eggs and the initiation of larval rearing trials. The occurrence of only a small number of spermiating males and ovulated females in the present study, compared to GnRHa-implanted individuals underscores the necessity and importance of the developed hormone induction method, in accelerating the efforts at controlling reproduction of bluefin tuna in captivity and in closing the cycle of production, thus removing pressure from the wild populations. The collection of eggs from the cages was extremely difficult, and it is certain that this will be a major bottleneck in any efforts at obtaining eggs from broodfish maintained in sea cages. The alternative would be to maintain bluefin tuna in land-based facilities, as it is the situation in yellowfin tuna (Wexler et al., 23). Such facility is now under construction by a private company (Clean Seas, personal communication). The development of land-based facilities for the maintenance of bluefin tuna in captivity will ensure proper collection of all eggs produced, either in response to GnRHa-induced or spontaneous spawning. More importantly, it will allow absolute control of all environmental parameters influencing reproductive function, especially water temperature which is an important cue regulating reproductive maturation. Finally, it will allow easier administration of the GnRHa implants, as the water of the tank can be reduced to whatever level is necessary for the implantation 171

172 5.3 EFFECT OF GNRHA TREATMENT ON THE REPRODUCTIVE ENDOCRINE AXIS (BRAIN- PITUITARY-GONADS) Introduction Earlier studies of the bluefin tuna reproduction pointed out June-July as a natural spawning period across the Mediterranean Sea (Medina, 22; Abascal, 23; Corriero, 23; Chapter 3). In addition, comparison of reproductive characteristics of wild and captive bluefin tuna breeders indicated that LH, the maturationalgonadotropin, accumulates normally in the pituitaries of captive bluefin tuna, and that the reproductive endocrine system in these fish is functioning and preparing for the natural spawning season (Chapter 4). These findings suggested that captive bluefin tuna, that do not complete spontaneously the processes of FOM, ovulation and spawning, could be induced to spawn by administration with exogenous GnRHa. This notion relays on vast experience with various cultured fish, which underscores high coincidence of successful spawning following GnRHa-based treatments, as long as the LH content in the pituitary is not a limiting factor (reviewed by Mylonas and Zohar, 21; Zohar and Mylonas, 21), and has been shown to hold true in bluefin tuna as well (Chapter 5.1). The current study examines the effects of GnRHa-implantation on key hormones along the reproductive axis of captive bluefin tuna broodstocks, such as native GnRH forms, LH and sex steroids Materials and Methods Experimental animals and husbandry Fish used in the experiments were caught around the Balearic Islands, Spain, by fishing campaigns that were carried out in 22 and 23 during the months May and June. A detailed description of the fish maintenance, experimental design, and specifications of the GnRHa-implants that were developed and used, are reported in sub-section Briefly, the first experiment included 32 fish held in a 5-m diameter cage. During mid-june 24, sixteen individuals of the latter population were treated with GnRHa implants at effective doses ranging between 4 to 8 µg GnRH kg -1 body weight (BW). The other 16 fish were kept in the same cage and remained untreated as controls. Each GnRHa implant was introduced to the fish in tandem with a tagging device that enabled both distinguishing the implanted fish and evaluating the implantation success. Five or 6 days after treatment, all fish were sampled for further hormone analysis. The second experiment included 4 fish held in two 25-m diameter cages (C 1 =12; C 2 =28). Three major modifications were introduced compared to the first-year experiment: (i) the GnRHa-implantation was delayed until early-july, during which sea water temperature was stabilized above 23 C, (ii) the GnRHa implants contained higher doses of GnRHa (5 1 µg GnRH Kg -1 BW), and their implantation was conducted in two runs spaced 4-days apart (first run: 3 th June, C 1 =12; second run: 4 th July, C 2 =19), (iii) the experimental fish were sampled in a time-course manner, i.e. 2 (n=14), 3 (n=14) and 8 (n=12) days after treatment Sampling procedures To evaluate the affects of GnRH-implantation on the reproductive endocrine axis the experimental fish were sampled 6 days after the first induction trial (24) and 2, 3 and 8 days after the second induction trial (25). One fish at a time was sacrificed using an underwater shot-gun. The fish was lifted onto a deck of a service boat. Morphometric parameters, such as fork length, wet BW and the weight of the gonads were recorded, and sex was determined by macroscopic observation of the gonads. Blood was sampled from the branchial arches, filled into tubes treated with heparin/ Pefabloc SC Plus (Roche, Mannheim, Germany) solution, and placed into an ice box. The plasma was separated by centrifugation and stored at -2 C until assayed. The head of each fish was severed onboard, placed on ice and transferred to the laboratory. There, the pituitary and brain were collected into separated tubes, frozen immediately in liquid nitrogen, and stored at -8 C until assayed. For histological evaluation of maturation stage, ovarian and testicular fragments were fixed in 4% formaldehyde, 1% gluteraldehyde buffered saline (McDowell and Trump, 1976) and embedded in glycol methacrylate resin (Technovit 71; Heraeus, Kulzer, Germany). Serial sections were obtained at a thickness of 3 µm on a microtome (Biocut 235, Reichert Jung, Germany) and were stained with methylene blue/ azure II/ basic fuchsin (Bennett et al., 1976) Hormone Analysis 172

173 Prior to sex steroid measurements, plasma samples were thawed and the steroids extracted twice with dichloromethane. For detection and measurement of 17-estradiol (E2), 17α,2β-dihydroxy-4-pregnen-3- one (17,2β -P), Testosterone (T) and 11-ketotestosteone (11-KT) a competitive ELISA procedure elaborated for other fishes (Cuisset et al., 1994) and modified by Nash et al. (2) and Susca et al. (21) was used. The pituitary and plasma LH content was measured using an ELISA developed to measure striped bass LH (Mañanós et al, 1997) modified and validated for tuna (Rosenfeld et al., 23; Chapter 4). The pituitary content of bluefin tuna native GnRH forms was measured using ELISAs previously developed for the quantification of cgnrh- II, sgnrh, and sbgnrh isoforms (Holland et al., 1998) Statistical Analysis Data was analyzed using the JMP IN 5.1 statistical software (SAS Institute Inc., Cary, USA). One-Way ANOVA was employed to compare mean values followed by Tukey-Kramer HDS (α =.5). Parameters in figures and text are presented as means ± SEM Results The effects of GnRHa-implantation on pituitary LH release During the first spawning induction trial (24), the LH levels in treated females 5-6 days post-treatment decreased significantly (P <.1) in the pituitary and increased (P <.5) in the plasma compared to the respective levels in the control (Fig. 5.9). Similar LH patterns were seen in GnRHa treated males, however, the pituitary LH content was almost 2-fold lower (P <.1) in males compared to females, while plasma LH levels did not vary significantly (P >.5) between sexes, neither in basal nor in GnRHa-induced conditions. Females Males Pituitary LH (mg/pit) ** 17.5 Control 15. GnRHa * Pituitary LH (mg/pit) 5 Control Females GnRHa Control GnRHa 75 Control Males GnRHa 75 Plasma LH (ng ml -1 ) * 5 25 *** 5 25 Plasma LH (ng ml -1 ) Control Females GnRHa Control Males GnRHa Fig. 5.9 Effects of GnRHa implantation on LH levels in the pituitaries (upper panel) and plasma (lower panel) of bluefin tuna females (left panel) and males (right panel) during the first spawning induction trial (24). Mean ± SEM levels in the pituitary and plasma are expressed as mg pituitary -1 and ng ml -1, respectively. Asterisks denote mean which were significantly different (* P <.5; ** P <.1; *** P <.1). In the second spawning induction trial (25), GnRHa treated females exhibited a gradual decrease in the content of pituitary LH throughout the experiment, and by day 8 post treatment ended up with less than 25% (11.79 ± 1.96 mg pituitary -1 ) of the initial LH content (Fig. 5.1). In males, consistent with the 24 results, the initial pituitary LH content was significantly (P <.5) lower compared to females. Although we missed out the pituitaries of treated males that were sampled at day 2 post treatment (were damaged upon 173

174 sacrificing the fish), it appears that the overall LH concentrations in the males' pituitary vary more rapidly compared to females. As early as day 3 post treatment the measured LH concentrations (2.45 ±.69 mg pituitary -1 ) were less than 2% of the initial pituitary LH content, whereas at day 8 post treatment the LH concentrations recovered (5.84 ± 1.18 mg pituitary -1 ) to approximately 5% of the initial LH content. In both sexes the treatment effect on plasma LH concentrations was the greatest at day 2 post treatment (Fig. 5.1). Yet, the latter LH peak (av ± ng ml -1 ) was statistically significant (P <.5) only in males. From day 3 to day 8 (end of the experiment), treated fish of both sexes had similar elevated (P <.1) plasma LH levels (av ± 9.93 ng ml -1 ) compared to the control (av ±.95 ng ml -1 ). Females Males pituitary LH (mg/pit) * 1 5 *** ** 1 5 Pituitary LH (mg/pit) C 2D 3D 8D C 3D 8D plasma LH (ng ml -1 ) * plasma LH (ng ml -1 ) C 2D 3D 8D C 2D 3D 8D Fig. 5.1 Effects of GnRHa implantation on LH levels in the pituitaries (upper panel) and plasma (lower panel) of bluefin tuna females (left panel) and males (right panel) during the second spawning induction trial (25). GnRHa-treated fish were sampled at days 2, 3, and 8 post treatment (2D, 3D and 8D, respectively). Mean ± SEM levels in the pituitary and plasma are expressed as mg pituitary -1 and ng ml -1, respectively. Asterisks denote means which were significantly different from Controls (* P <.5; ** P <.1; *** P <.1). Comparing data from the two spawning induction trials, distinguishes the 25 brooders with higher (P <.1) BW (av ± 4.38 kg) and lower (P <.1) gonado-somatic-index (GSI) values (1.45 ±.12) compared to 24 brooders (Table 5.2). Nevertheless, the plasma LH levels did not vary significantly (P >.5) in the 24 and 25 populations, neither in basal conditions (control groups), nor in GnRHastimulated conditions. Table 5.2 Comparison of some reproductive parameters of captive bluefin tuna broodstock. Basal plasma LH (ng/ml) GnRHa-stimulated plasma LH (ng/ml) BW (kg) GSI ± 2.47 (n=17) ± 11.8 (n=8) a ± 4.61 (n=27) 1.96 a ±.14 (n=27) ± 2.8 (n=13) ± 8.69 (n=13) b ± 4.38 (n=4) 1.45 b ±.12 (n=4) 174

175 Correlation between LH levels and successful GnRHa-implantation The pattern of pituitary vs. plasma LH in individual fish was used to confirm the occurrence of successful implantation. There was a good correlation between diver's indication of successful implantation (24 experiment) and the occurrence of pituitary LH release (Fig. 5.11). Female ZZ122 seems to be an exception, inasmuch as the tag indicates a good implantation (+), whereas the LH profiles resembled those of the controls, i.e. relatively high levels in the pituitary (av ± 9.34 mg pituitary -1 ) and low in the plasma (av ± 3.12 ng ml -1 ). In the second trial (25) no mismatches between diver's indication of successful implantation and the occurrence of pituitary LH release were found. All succesfully implanted fish, exhibited an increased plasma LH levels compared to the controls, whereas in badly implanted fish (i.e. fish ZZ214, ZZ219, ZZ226 and ZZ238) plasma LH levels did not vary significantly compared to the controls (Fig. 5.12). Females Pituitary LH (mg/pit) Plasma LH (ng ml -1 ) ZZ11 (+) ZZ113 (+) ZZ114 (-) ZZ119 (+) ZZ121 (+) ZZ122 (+) ZZ123 (+) ZZ125 (+) Pituitary LH (mg/pit) Males Plasma LH (ng ml -1 ) ZZ12 (-) ZZ14 (-) ZZ111 (-) ZZ112 Fig Correlation between the diver's indication of successful implantation and the occurrence of LH release (June 24). The pituitary (left axis) and plasma (right axis) LH profiles of GnRHa- treated females (upper panel) and males (lower panel) are given individually, concomitant with the respective diver's indication of successful implantation ( = failed implantation, + = successful implantation). (+) ZZ117 (+) ZZ12 (+) 175

176 Plasma LH (ng/ml) ZZ214 (-) 2D (+) zz219 (-) zz226 (-) 3D (+) zz238 (-) 8D (+) C Fig Correlation between the diver's indication of successful implantation and the occurrence of LH release (July 25). The plasma LH profiles of selected GnRHa- treated bluefin tuna specimens (i.e., ZZ214, ZZ219, ZZ226 and ZZ238), all exhibiting unsuccessful implantation, were compared to the respective mean value of the controls (C), and successfully implanted fish that were sampled at 2, 3 or 8 days post treatment (D2 (+); D3 (+) and D8 (+), respectively) The effects of GnRHa-implantation on pituitary content of native GnRH forms The levels of native GnRH forms (sgnrh, sbgnrh, cgnrh-ii) were measured in the pituitary extract of control and GnRHa treated fish. In both years, it appears that the GnRHa-implantation induced a gradual decrease in the pituitary content of sbgnrh (Fig. 5.13). This trend is more pronounced in males than in females, and becomes significant (P <.1) in males at day 8 days post treatment. No significant changes were found in the pituitary levels of cgnrh-ii, which were 1-fold lower compared o those of sbgnrh. The sgnrh levels were below the detection limit of our assay (<.5 pg/well). Fig Effects of GnRHa-EVAc implantation on sbgnrh content in the pituitaries of BFT during the first (left panel) and second (right panel) spawning induction trials. GnRHa-treated fish sampled at days 2, 3, 6 or 8 post treatment are marked as 2D, 3D, 6D, and 8D, respectively. Levels (mean ±SEM) are expressed as total amount (ng) per pituitary. Asterisks denote values significantly different from that of the control (C) fish (** P <.1). 176

177 The effects of GnRHa-implantation on plasma sex steroid profiles In 24, there were significant differences in steroid hormone levels between GnRHa-implanted and control fish (Fig. 5.14). In females, both T and E2 were significantly (P <.1) higher in GnRHa-implanted fish, whereas in males only 11-KT was significantly higher (P <.5). In 25, plasma T in GnRHa-treated females reached maximal levels (av ± 4.47 ng ml -1 ) 2 days post treatment, and were maintained about 3-fold higher (P <.5) compared to the control, also at day 3 post treatment (Fig. 5.15). Nevertheless, no significant (P >.5) change in female plasma E2 levels was detected through the experiment. In GnRHa treated males, significantly (P <.5) elevated plasma 11-KT levels were detected at day 2 post treatment and were maintained high up to day 8 post treatment. Although plasma T levels in GnRHa-treated males did not vary significantly from the control (P >.5) on the whole, their profile seems to resemble the respective profile of 11-KT (Fig. 5.15). At day 8 post treatment, the plasma 17,2P levels of both sexes exhibited a significant (P <.5) elevation (av ± 1.13 ng ml -1 ) compared to the control (av ±.58 ng ml -1 ). However in the other sampling points carried out at days 2 and 3 post treatment, the plasma 17,2P levels in GnRHa-treated fish did not vary significantly (P >.5) from the controls. Control 4 2 GnRHa 4 Steroids (ng ml -1 ) ** 1 ** * Steroids (ng ml -1 ) Testosterone Females Estradiol Testosterone Males 11-ketoT Fig Effects of GnRHa implantation on sex steroids plasma levels in bluefin tuna females (left panel) and males (right panel) during the first spawning induction trial (June 24). Levels (mean ±SEM) are expressed as ng ml -1. Asterisks denote values significantly different from that of the control (C) fish (* P <.5; ** P <.1). 177

178 Females Males Testosterone (ng ml -1 ) ** * Testosterone (ng ml -1 ) C 2D 3D 8D C 2D 3D 8D Estradiol (ng ml -1 ) * * KT (ng ml -1 ) 17,2βP (ng ml -1 ) C 2D 3D 8D 6 * C 2D 3D 8D 1. C 2D 3D 8D C 2D 3D 8D * ,2βP (ng ml -1 ) Fig Effects of GnRHa implantation on sex steroids plasma levels in bluefin tuna females (left panel) and males (right panel) during the second spawning induction trial (July 25). Levels (mean ±SEM) are expressed as ng ml -1. Asterisks denote values significantly different from that of the control (C) fish (* P <.5; ** P <.1) Ovarian and testicular development A detailed description of ovarian and testicular development of the 24 and 25 populations is given in the following sub-chapter (Chapter 5.3). Briefly, the results of the time-course study (25), classified two maturing groups among the GnRHa-treated females, whereas the control females had only late vitellogenic oocytes, which in most cases showed high atresia rates (Fig. 5.16a). The first group consisted of females sampled at day 8 post treatment (n=4), all exhibiting post-vitellogenic oocytes (Fig. 5.16b). The second group consisted of females sampled at day 2 and 3 post treatment, all exhibiting the advanced stages of final oocyte maturation (Fig. 5.16c). It is important to note that these fish were hormonally treated at the same day (4 th July), maintained in the same cage, but were sampled at two consecutive days (6 th and 7 th July). a b c GV GV GV GV GV Fig Histological appearance of BFT post-vitellogenic oocytes during the process of maturation. (a) Section typifying control females with atretic follicles (marked by an arrowhead), and oocytes exhibiting the nucleus (germinal vesicle, GV) at a central position. (b) Section typifying females 8 days post GnRHa treatment. The most advanced oocytes exhibit migratory GV. (c) Section typifying females 2-3 days post GnRHa treatment. The yolk granules within the most advanced oocytes have completed coalescence and GV is at the periphery 178

179 5.3.4 Discussion This study focused on the effects of GnRHa-based spawning induction treatments on key hormones along the brain-pituitary-gonad axis of captive bluefin tuna. Our first spawning induction trial was carried out during mid-june (24) known to be the natural spawning period of bluefin tuna across the Mediterranean Sea (Medina, 22; Abascal, 23; Corriero, 23; Chapter 3). Five or 6 days after the GnRHa-implantation a significant decrease in the levels of pituitary LH, with simultaneous increased LH levels in the plasma of treated fish vs. the controls, indicated a positive effect of the GnRHa treatment. A good correlation was found between the occurrence of LH release in individual fish and the respective diver's notification as to the success of the GnRHa implantation procedure. These results reinforced the reliability of the newly developed GnRHa-delivery device, which enables a simultaneous implantation and tagging, thereby minimizes the occurrence of stress among treated BFT. Triggered by the high plasma levels of LH, the sex steroid profiles fluctuated, with treated females demonstrating elevation in plasma levels of T and E2, and treated males demonstrating elevation in plasma levels of 11-KT. The positive effects of the treatment were confirmed also by histological analysis of ovarian fragments that showed high frequency of oocytes at the more advanced developmental stages, i.e. germinal vesicle (GV) migration, GV breakdown (GVBD), and low frequency of atresia among GnRHa-treated females compared to the controls (See also next Chapter 5.3). Nevertheless, the first trial (24) ended up with a partial success only, as cage spawning could not verified, neither with monitoring spawning behavior nor with the collection of fertilized eggs within the cages (see Chapter 5.1). Given that the success of the GnRHa-treatment is largely dependent on its precise matching with the completion of vitellogenesis (Mylonas et al., 1997c), an effort was made during the second trial (25) to time the treatment better with the ovarian post-vitellogenic stage. Accordingly, the second spawning induction trial was delayed until the first week of July in the following year. During that period the sea surface temperature at the rearing location (La Azohia, Spain) stabilized over 23 C, which is thought be a threshold for gonadal maturation in several tuna species, including the bluefin tuna (Schaefer, 21; see also Chapter 3). Additionally, in order to increase our chances to induce maturation and spawning, the GnRHa implantation was conducted in two runs spaced 4 days apart. The time-course study indicated a rapid increase of plasma LH at 2 days post treatment, and thereafter elevated circulating levels up to day 8 when the experiment was ended. The plasma LH profiles paralleled the GnRHa kinetic release from the EVAc implants (see Chapter 5.1). Similar effects of sustained GnRHa-delivery systems on LH profiles were seen in the gilthead seabream (Sparus aurata) (Zohar et al., 199; Zohar, 1996), a non-synchronous spawner like the bluefin tuna (Zohar et al., 1995). Normally, gilthead seabream females exhibit a characteristic daily fluctuation of plasma LH levels, with a peak 8 h before ovulation and a dramatic decrease 4 h after ovulation (Gothilf et al., 1997). It is worth mentioning that even under GnRHa-induced conditions, whereby sustained elevated plasma LH levels override the natural daily fluctuation of plasma LH, the seabream females manage to cope with their typical daily cycle of FOM, ovulation and spawning (Zohar et al., 1995). Nevertheless, in synchronous spawners like striped bass (Morone saxatilis) (Mylonas et al., 1997c), white bass (Morone chrysops) (Mylonas et al., 1998b), and rainbow trout (Oncorhynchus mykiss) (Breton et al., 199), the GnRHa treatments generate different patterns of plasma LH. In these species, as long as high GnRHa levels are being supplemented by the controlled delivery system, the plasma LH levels gradually increase and then show a sharp rise upon ovulation. Such a pattern also characterizes the natural plasma LH levels in wild populations of striped bass (Mylonas et al., 1997c; Hassin et al., 1998), and many other synchronous spawners including salmonid species (Prat et al., 1996, Breton et al., 1998, Davies et al., 1999, Gomez et al., 1999, Santos et al., 21), African catfish (Schulz et al., 1997), black carp (Gur et al., 2), and the European sea bass (Dicentrarchus labrax) (Mateos et al., 23). The mechanisms underlining dissimilar responses of the pituitary to GnRHa stimuli in non-synchronous vs. synchronous spawners, are still unknown. However, an ultimate mode for such regulation could be the GnRH receptors (GnRH-Rs), expressed on the pituitary gonadotropes, whos density is likely to determine the ability of the cells to respond to GnRH. Moreover, findings in masu salmon (O. masu), indicate the existence of multiple genes for GnRH- Rs (Jodo et al., 23), which are being expressed differentially during the reproductive season (Jodo et al., 25). The GnRH-R gene multiplicity appears to be a common phenomenon among teleosts (Bogerd et al., 22; Illing et al., 1999; Millar et al., 21, 24; Okubo et al., 21; Wang et al., 21; Parhar et al., 25; Moles et al., 26), as well as among other vertebrates including humans. Therefore, we tend to think that such a system that exhibit quantitative and qualitative changes throughout the reproductive season could also provide a good platform for directing species specific gonadotropic responses to GnRH. Consistent with our previous results (see Chapter 4) the LH content within the pituitaries of bluefin tuna brooders ranged from one to a few tens of milligrams per pituitary gland, which is about 2-3 orders of magnitude higher compared to the respective levels in smaller perciformes, like the gilthead seabream (Holland et al., 1998) and striped bass (Hassin et al., 1998). Nevertheless, all examined species, share 179

180 similar plasma LH levels (few to tens ng per ml plasma). Thus, the remarkably high pituitary LH content found in bluefin tuna seems to be vital to make up for high dilution rates when released into the circulation that maintains the large body size of this fish. Our findings demonstrate a sexual dimorphic pattern of the pituitary LH, with females accumulating nearly 2- to 3-fold higher levels compared to males. With the limitation of a single time-course study, it appears that the release kinetics of LH from the pituitary storage also differ between the sexes. For example, stimulated by GnRHa, the pituitary LH content in females gradually decreased throughout the experiment (8 days post treatment), whereas in males, shortly after treatment (2-3 days) the pituitary LH levels dramatically decreased (lost more than 8% of the initial concentrations) and were already recovering at day 8 post treatment. Studies that profiled the pituitary LH content (Zohar, unpublished), and the expression levels of the gene encoding for the LH -subunit (Elizur et al., 1995; Meiri et al., 24) throughout the natural spawning season of gilthead seabream, demonstrated a sexual dimorphic pattern that resembled the one found in bluefin tuna. Although the exact mechanism reinforcing sex related patterns of pituitary LH is still unknown, two lines of evidence suggest a role for the dynamic GnRH / GnRH-Rs system. The first evidence refers to a study with masu salmon showing sex related differences in the regulation of the GnRH-Rs genes by GnRHa, exemplified by their induced expression in females and inhibited or unchangeable expression in males (Jodo et al., 25). The second evidence refers to sexual differences in (i) size and number of GnRHimmunoreactive cells among the preoptic area (POA), with males encompassing larger and more numerous cells compared to females (Grober and Bass, 1991; Rissman and Xia, 1998; Elofsson et al, 1999; Ishizaki et al., 24) and (ii) synthesis and releasing activities of the POA-GnRH neurons, which mostly synthesize sbgnrh (Amano et al., 1994; Ishizaki et al., 24). Interestingly, our study with bluefin tuna reveals no apparent sexual differences in the pituitary content of sbgnrh, yet its releasing activities following GnRHa implantation were more pronounced in males than in females, with the formers exhibiting a significant loss at day 8-post treatment. The GnRHa implantation was found to reduce the rates of follicular atresia among treated BFT females compared to the controls. Generally, atresia, the programmed cell death process (apoptosis) within the follicles, is associated with the absence of pituitary LH surge (Hunter et al., 1986; Schaefer, 1998; Mylonas and Zohar, 21). Accordingly, the sustained elevated LH levels in the plasma of GnRHa bluefin tuna in both trials seem to help preserving the viability of post-vitellogenic oocytes that form the next batch to mature. This notion is further supported by our recent findings showing that physiological levels (5-1 ng ml -1 ) of recombinant gilthead seabream LH attenuate significantly the occurrence of atresia among cultured post-vitellogenic follicles compared to untreated ones (Meiri and Rosenfeld, unpublished). As was suggested by the mammalian model, the atretic degeneration could be suppressed directly by LH, or indirectly by paracrine factors, considered to mediate the LH signals from the somatic cells toward the oocyte (reviewed by Kaipia and Hsueh, 1997; Markstrom et al., 22; Rolaki et al., 25). These factors include insulin-like growth factor-i and II (IGF-I and IGF-II), epidermal growth factor (EGF) and members of the transforming growth factor (TGF ) super-family (reviewed by Ge, 25). The pituitary LH surge that normally accompanies the completion of vitellogenesis, triggers a switch in the steroidogenic activity of the follicular somatic cell layers, giving rise to a rapid decline in plasma E2, and a dramatic elevation in the plasma levels of the maturation inducing steroid (MIS; Nagahama, 1994, Nagahama et al., 1994, Selman et al., 1994). In turn, the MIS acts at the level of the oocyte membrane to trigger a cascade of intracellular events leading to the resumption of meiosis and completion of oocyte maturation (Peter and Yu, 1997). The noticeable changes along this process include GV migration towards the animal pole, GVBD, yolk clarification and an increase of oocyte volume due to hydration. Among the positively recognized MISs the most common are the 17,2β-P and 172β,21-trihydroxy-4-pregnen-3-one (2-S) (Scott and Canario, 1987). Coincidental findings in the current study highlight 17,2β-P as a likely MIS in BFT. In that respect, at day 8-post treatment the GnRHa-implanted BFT females and males exhibited remarkably elevated plasma 17,2β-P levels, which were approximately 3-fold higher compared to the controls. The females among this group were highly synchronized and showed post-vitellogenic oocytes with migratory GV, as the maximal developmental stage. The females sampled at day 2 and 3 days post treatment also exhibited a synchronized gonadal development, however, their most advanced oocytes were at the GVBD/hydration stage. The plasma 17,2β-P levels of the latter females did not vary significantly from those of the controls. A detailed study, which trailed a daily spawning cycle in the gilthead seabream females, shows similar matching between the morphological characteristics of the most advanced oocyte population within the ovary, and the plasma levels of 17,2β-P (Gothilf et al., 1997). Accordingly, the 17,2β-P plasma levels increase parallel to GV migration, and peaked concomitant with the preovulatory LH surge once GV arrives the periphery. The rise in plasma 17,2β-P levels was relatively short (from 12 to 8 h before spawning), and by the time of GVBD (4-h before spawning), a decline to basal levels was observed. Thus, considering that gilthead seabream and bluefin tuna share similar spawning intervals (av. of 1.1 days; Medina, 22; see also Chapter 5.1), one may anticipate that the bluefin tuna sampled at day 2 and 3 post 18

181 treatment were about 4 h before spawning, whereas those sampled at day 8 post treatment would have spawned 12 to 8-h relatively to their sampling time. Our results indicate that the GnRHa treatments have helped synchronizing gonadal development within and between the sexes of bluefin tuna held in the same cage. However, females in the neighboring cages appeared to be shifted in respect to their spawning time, although the implantation procedures for both cages were conducted at an equivalent timing (9:3-12: AM). It is worth mentioning that such division to "early" and "delayed" spawners (spawning time at 7: AM and 3: PM, respectively) was recognized among gilthead seabream that were reared at similar condition but in two separated tanks (I. Meiri, unpublished data). In conclusion, this study clearly demonstrates a positive effect of the GnRHa implantation on the reproductive brain-pituitary-gonad axis of captive bluefin tuna, and draws attention to a striking resemblance in reproductive traits of this fish and other asynchronous spawners such as the gilthead seabream. The more noticeable similarities include the LH response to sustained release of GnRHa, and the remarkable association of the plasma 17,2β-P levels and the maturation stage of the oocytes during a daily spawning cycle. On the basis of the plasma 17,2β-P elevation in both BFT sexes at the time of GV migration, we suggest a role for this steroid in inducing FOM and ovulation. Nevertheless, further studies are required to positively confirm the identity of bluefin tuna MIS. Another future research should verify whether the aforementioned similarities between bluefin tuna and the gilthead seabream, are attributed to the fact that both species are non-synchronous spawners that perform daily spawning cycles. 5.4 EFFECT OF GNRHA TREATMENT ON GONADAL MATURATION BASED ON HISTOLOGICAL ANALYSIS Introduction The bluefin tuna ovary resembles that of other teleost fish, consisting of ovigerous lamellae with numerous follicles at different stages of development (Corriero et al., 23). Each developing oocyte is surrounded by a single layer of follicular cells. Seven oocyte developmental stages have been described: perinucleolar (diameter µm), lipid (diameter µm), early vitellogenesis (diameter 22-3 µm), late vitellogenesis (3-5 µm), migratory nucleus (5-65 µm), pre-hydrated (65-75 µm) and hydrated (75-9 µm). The simultaneous presence of all of the oocyte developmental stages in fish during the spawning period (Medina et al., 22; Corriero et al., 23) indicates that bluefin tuna has an asynchronous oocyte development and is a multiple spawner (Tyler and Sumpter, 1996). The bluefin tuna testis is constituted by seminipherous tubules radiating from the longitudinal main sperm duct toward the periphery (Abascal et al., 23). The testicular structure is cystic: each cyst contains germinal cells in the same development stage, branched by the cytoplasm of somatic cells (Sertoli cells). The presence of spermatogonia along the greater part of the testicular tubules indicates that the bluefin tuna testis is of the unrestricted spermatogonial type according to the classification of Grier et al. (198). The bluefin tuna reproductive cycle in the western and central Mediterranean Sea has been described on the basis of the histological appearance of the ovary (Corriero et al., 23; Zubani et al., 23) and testis (Santamaria et al., 23). Fish were reproductively inactive from August to March, when only perinucleolarstage oocytes were present in the ovaries, while the testes had germinal cysts containing mainly spermatogonia and spermatocysts. Gonad development started in April when oocytes at lipid stage appeared in the ovaries while all the spermatogenic stages were present in the seminipherous tubules. During May ovaries were characterized by the presence of vitellogenenic oocytes, while in the testes the lumen of seminipherous tubules filled progressively with spermatozoa. Gonads in spawning condition (ovaries with hydrated oocytes and/or post-ovulatory follicles and testes replete with spermatozoa) were found in late June to early July. During late June-September, ovaries showed only perinucleolar-stage oocytes, as well as late stages of atresia of vitellogenic follicles; only residual spermatozoa were present in the testes. Natural spawning of bluefin tuna occurs throughout the Mediterranean Sea, including the Balearic Islands (western Mediterranean) (Nishida et al., 1997; Susca et al., 21; Medina et al., 22; Corriero et al., 23), Malta and the South Tyrrhenian Sea (western Mediterranean) (Nishida et al., 1997) and the Levantine Sea (eastern Mediterranean) (Karakulak et al., 24) (Oray and Karakulak, 25). In the Levantine Sea reproduction occurs in May, almost one month before the other spawning areas, because sea surface temperature suitable for bluefin reproduction (over 22 C) are reached earlier than in the central and western Mediterranean (Karakulak et al., 24). 181

182 The objectives of the present study were to (a) examine the potential for maturation of wild-caught Mediterranean bluefin tuna kept for a period of 1-3 years under culture conditions and (b) study the effect of the GnRHa treatment on reproductive maturation based on evaluation of gonadal histology Materials and Methods Gonads from a total of 66 captive-reared bluefin tuna specimens were used. The origin of the fish, their husbandry, and the preparation and validation of the GnRHa-delivery system and GnRHa implantation procedures was described in the previous Chapter (5.1). In fish were used, of which 17 (7 females and 1 males) were untreated (Control) and 9 (6 females and 3 males) were treated with GnRHa (Table 5.1). For the 25 experiments, 4 fish were used, 14 of which were Control (7 females and 7 males) and 26 (1 females and 16 males) were implanted with GnRHa. In 24 all the fish were sacrificed 5-6 days after GnRH implantation, while in 25 they were killed after 2-3 or 8 days after treatment (Table 5.1). From each fish fork length (FL) was measured to the nearest cm, total body weight (W B ) and gonad weight (W G ) were measured at the nearest kg and 1 g, respectively, and the gonadosomatic index (GSI) was calculated as 1 W G W B - 1. Gonads were dissected onboard, placed on ice in a styrofoam box and transported to the laboratory onshore within 1 h. Gonad slices were fixed in Bouin s fixative or 1% buffered formalin, dehydrated in increasing ethanol concentration, clarified in xylene and embedded in paraffin wax. Five-µm thick sections were cut and stained with Haematoxylin-Eosin. For the classification of reproductive state of females, the most advanced oocyte stage was recorded for each specimen, according to the classification of Corriero et al. (23), with the exception that the category final maturation stage was used to identify oocytes at migratory nucleus and following stages. The presence of atretic and postovulatory follicles was also recorded. Vitellogenic atretic follicles were classified at α or β stage according to Hunter et al. (Hunter et al., 1986). Briefly, the identification of α atretic follicles was made on the basis of zona radiata fragmentation, yolk granule break down and reabsorbtion, as well as nuclear disarrangement. In β atretic follicles, yolk was completely reabsorbed, germinal vesicle disappeared and oocytes were invaded by follicular and theca cells. For the classification of reproductive state of males, the type of spermatogenic cysts was recorded and the quantity of spermatozoa in the lumen of seminipherous tubules was subjectively evaluated. In order to verify if a normal yolk accumulation and vitellogenic oocyte growth occurred in fish reared in captivity, the diameters of fully vitellogenic oocytes, the surface occupied by eosinophilic yolk granules, as well as the diameter of eosinophilic yolk granules were compared between wild and captive-reared fish. For this aim, sections of 3 fully vitellogenic oocytes having a large and centrally located nucleus were selected from 6 captive-reared fish and compared with an equal number oocytes at the same stage from a total of 6 wild adult fish caught in the same period around Malta waters. Oocyte diameter and oocyte surface occupied by yolk granules were measured from microphotographs taken with a digital camera (DC 3, Leica, Cambridge, U.K.) connected to alight microscope (DMRBE, Leica, Cambridge, U.K), using an image analysis software (QWIN, Leica, Cambridge, U.K.). Estimations of the number of oocytes of different maturation stage contained in the ovaries (including atretic follicles) were carried out by means of the stereological method of Weibel and Gómez (1962). Stereometry permits the calculation of the number of oocytes per unit volume of ovary (NV, numerical density) from microscope images, according to the formula: 3/ 2 K N A NV = 1/ 2 β VV where β is a shape coefficient, K is a size distribution coefficient, NA is the number of oocyte transections per unit area, and VV is the partial area of oocytes of a given category (volume fraction or volume density). From digital micrographs of histological sections, VV was determined by image analysis of digital micrographs using the software Image J, NIH, USA. The total number of oocytes of each category contained in the gonad of each individual was obtained by multiplying NV by the entire ovarian volume (Medina et al., 22). Oocyte stereological quantification was carried out for 29 captive-reared bluefin tuna (13 untreated Control and 16 GnRHa treated fish) and, for comparison, for 7 wild adult specimens captured by purse seiner during the same period in the waters around the Balearic Islands. Data from oocyte diameters, surface occupied by yolk granules and oocyte quantification carried out by the stereological method, are expressed as means ± SEM and were analyzed by Student s t-test at minimum significance of p <

183 5.4.3 Results The size of the bluefin tuna ranged between 138 and 218 cm FL and, according to the size of first sexual maturity reported by Corriero et al. {, 25 #3224, all of them were adult Histological evaluation of stage of maturation in females (Table 5.3) Table 5.3 Identification number, biometric data and histological characteristics of the ovaries from bluefin tuna reared in captivity. All fish from 24 were sacrificed 5-6 days after GnRHa treatment; fish from 25 were sacrificed 2-3 days (a) or 8 (b) days after treatment. Identification number FL (cm) W B (kg) W G (g) GSI Treatment More advanced oocyte stage* Postovulatory follicles control late vitellogenesis no control late vitellogenesis* no control late vitellogenesis yes control late vitellogenesis no control late vitellogenesis no control late vitellogenesis* no control lipid no treated lipid** no treated final maturation yes treated final maturation yes treated final maturation yes treated late vitellogenesis yes treated final maturation yes control late vitellogenesis* no control late vitellogenesis* no control late vitellogenesis yes control late vitellogenesis* no control early vitellogenesis no control late vitellogenesis no control late vitellogenesis* no treated (a) final maturation yes treated (a) final maturation yes treated (a) final maturation yes treated (a) late vitellogenesis yes treated (a) late vitellogenesis* no treated (a) late vitellogenesis yes treated (b) late vitellogenesis yes treated (b) late vitellogenesis yes treated (b) late vitellogenesis yes treated (b) late vitellogenesis yes FL, fork lenght; W B, total body weight; W G, gonad weight; GSI, gonadosomatic index; *, major α atresia of late vitellogenic oocytes (>5% atresia of late vitellogenic oocytes); **, β atresia of late vitellogenic oocytes. The stage of final maturation included the migratory nucleus stage and the stage preceding oocyte hydration (stage of germinal vesicle breakdown). Among un-treated Control females in 24, 1 individual had the most advanced oocyte population at the lipid stage (Fig. 5.17a) and 6 individuals had oocytes at the late vitellogenesis stage (Fig. 5.17b). Among the latter fish, 2 displayed major α atresia (i.e. > 5% atresia of late vitellogenic oocytes) (Fig. 5.17c) and 1 showed postovulatory follicles. Among GnRHa-treated bluefin tuna, 1 had the lipid stage as the most advanced oocytes stage, but displayed also β atresia of vitellogenic oocytes (Fig d), 1 had late vitellogenic oocytes, while the presence of oocytes in final maturation was evidenced in all of the other 4 specimens (Fig. 5.17e). Postovulatory follicles were found in all GnRHa-treated fish except, the one devoid both of oocyte at late vitellogenesis and final maturation stage. In 25, 1 out of 7 Control fish had oocytes at early vitellogenesis as the most advanced development stage ((Fig. 5.17f). All the other Control specimens had late vitellogenic oocytes and, among them, one had also postovulatory follicles and 4 showed major α atresia. Among 1 GnRHa-treated individuals, 7 had late vitellogenic oocytes as the maximum oocyte stage and 3 showed also final oocyte maturation. Postovulatory follicles were observed in all GnRHa-treated fish (Fig. 5.17e), except in a late vitellogenic one that displayed also major α atresia. 183

184 Fig Micrographs of the ovaries from different bluefin tuna reared in captivity in floating cages off the coast of La Azohía (Murcia, Spain). The fish were sacrificed in 24 and 25 during the natural spawning season of the species in the Mediterranean. (a) Ovary from an untreated control fish showing the lipid stage as the most advanced oocyte stage. (b) Ovary from an untreatd control with oocytes at late vitellogenesis stage (lv). (c) Ovary from a control individual with oocytes at late vitellogenesis stage and major α atresia of vitellogenic follicles. (d) Ovary from a GnRHa-treated specimen with the most advanced oocytes at the lipid stage showing advanced atresia (β atresia) of vitellogenic follicles. (e) Oocytes at migratory nucleus stage togheter with post-ovulatory follicles in a GnRHatreated spawning specimen. (f) Ovary from a GnRHa-treated animal having oocytes at early vilettogenesis as the maximum oocyte stage. Haematoxylin-eosin staining. Bar = 3 µm in (a), (b), (d) and (f); 6 µm in (c) and (e). arrowhead, postovulatory follicle; arrow, β atresia of vitellogenic follicles; double arrow, a atresia of vitellogenic follicles; *, fragmented zona radiata in an atretic oocyte vitellogenic follicle; lv, oocyet at late vitellogensesis stage; mn, migratory nucleus stage oocyte. 184

185 Histological evaluation of stage of maturation in males (Table 5.4) Table 5.4 Identification number, biometric data and histological classification of the testes from bluefin tuna reared in captivity. All fish from 24 were sacrificed 5-6 days after GnRHa treatment; fish from 25 were sacrificed 2-3 days (a) or 8 (b) days after treatment. Identification FL W B W G GSI Treatment Histological Spermiating** number (cm) (kg) (g) classification* control late spermatogenesis no control late spermatogenesis no control early spermatogenesis no control late spermatogenesis yes control late spermatogenesis no control early spermatogenesis no control early spermatogenesis no control late spermatogenesis yes control late spermatogenesis no control spent no treated late spermatogenesis no treated early spermatogenesis no treated late spermatogenesis no control spent no control late spermatogenesis no control late spermatogenesis no control early spermatogenesis no control late spermatogenesis no control late spermatogenesis no control late spermatogenesis no treated (a) early spermatogenesis no , treated (a) late spermatogenesis no treated (a) early spermatogenesis no treated (a) late spermatogenesis no treated (a) late spermatogenesis no , treated (a) late spermatogenesis yes treated (a) late spermatogenesis yes treated (a) late spermatogenesis no treated (a) late spermatogenesis yes treated (a) late spermatogenesis yes treated (b) late spermatogenesis yes treated (b) late spermatogenesis no treated (b) early spermatogenesis no treated (b) late spermatogenesis no treated (b) late spermatogenesis no treated (b) late spermatogenesis no FL, fork lenght; W B, total body weight; W G, gonad weight; GSI, gonadosomatic index; *The histological classification was based on the types of spermatocysts observed in the germinal epithelium and the abundance of spermatozoa in the lumen of seminipherous tubules (further details are reported in the text). **Fish were classified as spermiating if they released semen either spontaneously when on board, or in response to a pressure on the abdomen. Among the 1 un-treated Control males in 24, 3 individuals were classified at early spermatogenesis as their testes showed most of the germinal epithelium occupied by cysts containing spermatocytes and spermatids, but cysts containing spermatozoa and luminal spermatozoa were also observed in most of seminiferous tubules (Fig Six individuals had testes that were classified at late spermatogenesis as their germinal epithelium consisted mainly of cysts containing spermatids and spermatozoa, and the lumen of seminipherous tubules was filled with spermatozoa (Fig Two fish of this last group were in spermiating conditions, i.e. sperm was released from the testes when the fish were brought on board, either spontaneously or in response to external pressure. One specimens was considered spent since the germinal epithelium consisted exclusively of spermatogonia, and only few residual spermatocysts were visible, mainly with spermatids or spermatozoa; luminal residual spermatozoa could still be observed (Fig. 5.18) Of the 3 individuals implanted with the GnRHa-implants, 1 of them was at early spermatogenesis and 2 at late spermatogenesis. In 25, among 7 untreated Controls, 1 was classified at early spermatogenesis, 5 at late spermatogenesis and 1 was spent. Of the 16 GnRHa-treated fish, 3 were at early spermatogenesis and 13 at the late spermatogenesis stage. Five GnRHa-implanted males with testes at the late spermatogenesis were also spermiating. 185

186 Fig Micrographs of the testes of different bluefin tuna reared in captivity in floating cages off the coast of La Azohía (Murcia, Spain). The fish were sacrificed in 24 and 25 during the natural spawning season of the species in the Mediterranean. (a) Testis from an untreated control fish classified at early spermatogenic stage showing germinal epithelium occupied by cysts containing spermatocytes and spermatids; luminal spermatozoa can also be observed. (b) Testis from a GnRHa treated individual classified at late spermatogenesis. The germinal epithelium consists mainly of cysts containing spermatids and spermatozoa and the lumen of seminipherous tubules is filled with spermatozoa. (d) Testis from a GnRHa treated spent bluefin tuna. The wall of seminipherous tubules is often devoid of spermatocysts and residual spermatozoa can be observed in the lumen of seminipherous tubules. Haematoxylin-eosin staining. Bar = 3 µm. arrow, germinal epithelium; *, wall of seminiperour tubules devoid of spermatocysts; l, lumen of seminipherous tubules; sz, spermatozoa Yolk quantification and stereological analysis The diameter of fully vitellogenic oocytes, the oocyte surface occupied by eosinophilic yolk and the diameter of yolk granules of captive-reared and wild bluefin tuna did not exhibit any significant differences (Table 5.5). No statistically significant difference was found in the volume fraction of atretic oocytes between GnRHatreated and wild fish, while a significantly higher quantity of atretic oocytes was evidenced in untreated Controls. The number of oocytes at the lipid and vitellogenic stages did not differ among the three groups. No oocytes at final maturation were found in the ovaries of captive untreated bluefin tuna, while GnRHatreated individuals had a significantly higher number of oocytes at final maturation than wild fish (Table 5.6). 186

187 Table 5.5 Diameter of fully vitellogenic oocytes, surface occupied by yolk and diameter of yolk granules for both captive-reared and wild bluefin tuna. Data are mean (± S.E.) of 3 measures from 6 different fish per group (captive-reared and wild). Oocyte diameter Yolk surface Yolk granule diameter (µm) (µm 2 ) (µm) Captive-reared 434.4± ± ±.2 Wild 435.1± ± ±.5 Table 5.6 Stereological data from ovarian histological sections of wild and captive reared bluefin tuna. Wild Captive (untreated) Captive (GnRHa-treated) (n = 7) (n = 13) (n = 16) Atretic oocytes.2±.4 a.15±.13 b.4±.5 a Vv Lipid vesicle 7.6±239.2 a ±553.6 a 1161±274.6 a (No g -1 W B) Vitellogenesis 245.5±145.2 a 139.3±113.9 a 13.7±74.4 a (No g -1 W B) Final maturation 8.96±11.4 a 24.8±25.1 b (No g -1 W B) Different superscripts indicate statistically significant differences Discussion The need to relieve the fishing pressure on the wild Atlantic bluefin tuna stock, without depriving humankind of an economically important, tasteful and highly nutritive feeding resource, has led the scientific community to attempt to close the life cycle of this fish in captivity {Lioka, 2 #273}(Doumenge, 1996). As already discussed. most fish reared in captivity exhibit some degree of reproductive dysfunction, presumptively due to the combination of several factors such as the lack of the appropriate "natural" spawning environment (Zohar, 1989a; Zohar, 1989b; Battaglene and Selosse, 1996), or even the lack of essential dietary components in the broodstock diet (Watanabe and Vassallo-Agius, 23). Captivity-induced reproductive impairments often result in failure of completion of oogenesis (Zohar, 1989b; Peter et al., 1993) or, for males, in the production of low quality-milt (Billard, 1986, 1989). The control of reproduction represents one of the most crucial step towards the objective of domesticating bluefin tuna (Sawada et al., 25). In teleost fish, as in other vertebrates, gametogenesis consists of mitotic proliferation, meiotic division, and differentiation of germ cells from diploid, primary germ cells to haploid, mature gametes (Redding and Patino, 1993). During oogenesis, meiosis is interrupted twice: during prophase I and during metaphase II. During the prophase I arrest, the oocytes undergo a remarkable increase in size caused by an enormous accumulation of yolk (Tyler and Sumpter, 1996; Brooks et al., 1997). Yolk accumulation is due to oocyte uptake of vitellogenin, a phospholipoprotein synthesized by the liver under the stimulation of the sex steroid hormone 17β-estradiol, transported in the bloodstream to the follicular layer and then endocytosed by oocytes (Wallace and Selman, 1985). Yolk content is an important determinant of egg quality in fish as it represents the major nutrient for the developing embryo (Brooks et al., 1997). Based on the histological evaluation of captive (this chapter) and wild females (earlier chapter), no differences attributed to the captive rearing, in terms of oocyte growth, were observed. For example, no difference was found between captive and wild individuals sampled in the same period, in the size of oocytes at the end of the vitellogenic growth. In addition, the surface occupied by yolk granules, as well as their size was not different between the two groups. Finally, no differences in terms of morphology and staining affinity of yolk granules were observed between captive and wild bluefin tuna (data not shown). These observations suggest that vitellogenin uptake and oocyte growth were not affected by the rearing of adult bluefin tuna collected from the wild in captivity for a period of 1-3 years. The results are encouraging as they indicate that it is possible to rear wild-caught, mature bluefin tuna under captive conditions, in this case coastal sea cages, and have them undergo reproductive maturation similar to wild individuals. In many captive fishes, once oocyte growth is completed, the lack of the appropriate luteinizing hormone (LH) levels can result in failure to undergo final oocyte maturation, and the oocytes undergo atresia, an irreversible form of programmed cell death (see review by Mylonas and Zohar, 21a). In the present study, all but two Control fish, had oocytes at late vitellogenesis at the time of sacrifice, but 5% of them showed major α atresia of vitellogenic follicles, a condition indicative of cessation of reproductive activity (Hunter et al., 1986; Schaefer, 1998). Oocytes at final maturation were not found in any of the Control fish, but two of them (one each from 24 and 25) contained postovulatory follicles in their gonads. It may be assumed that these two fish had spawned shortly before sampling, as postovulatory follicles in tunas can be distinguished up to 24 h after ovulation (Hunter et al., 1986). This finding seems to indicate that, after 1-3 years of life in captivity, bluefin tuna do have a certain capacity to bring oogenesis to an end and ovulate 187

188 spontaneously. Although in the present study, cage spawning could not be verified, either with recording of spawning behavior or collection of fertilized eggs from the vicinity of the cages, the obtained results underline that the possibility, then, remains that after some period (years?) of acclimation to the captive environment, and given the appropriate conditions, wild-caught, adult bluefin tune fish could spawn spontaneously and produce viable embryos. The timing of a hormonal treatment is essential for successful induction of ovulation, and such treatment must be done when the fish have completed vitellogenin uptake; if delayed further, GnRHa treatment can be unsuccessful due to atretic degeneration of the vitellogenic oocytes (Mylonas et al., 1997b). As described in the previous section, the GnRHa-implantation experiments in 24 and 25 were carried out during the natural spawning period of the species in the central-western Mediterranean. Especially during 25, the hormonal treatment was delayed until sea surface temperature was >23 C, which is the spawning temperature threshold in several tuna species (Schaefer, 21). Among GnRHa-treated bluefin tuna, 1 individual had ovaries containing β stage atresia and no vitellogenic follicles, indicating that oocyte apoptotic degeneration started several days before sampling and presumptively before the hormonal treatment. Another specimen had ovaries with major α atresia of vitellogenic follicles. All other implanted fish showed final oocyte maturation and/or postovulatory follicles, thus indicating that, at the time of sampling, they were going to ovulate or had just ovulated. Not considering the β atretic fish, the hormonal treatment was effective in 14 out of 15 female bluefin tuna, which is a 93% induction of ovulation success. Oocytes at final maturation occurred only in fish sacrificed 2-6 days after hormone administration, but not in fish sampled 8 days after implantation. This finding, although based on a small number of specimens, suggests a shortterm effect of the treatment and is in agreement with the reported plasma GnRHa levels, which were significantly elevated in GnRHa-implanted for 7 days after implantation. The spawning fraction of GnRHa-implanted fish (.93) can be extrapolated to spawning intervals of 1.1 days, which is comparable to that obtained from samples of wild spawners caught at the peak of the reproductive season (Medina et al., 22). Another effect of the GnRHa treatment was the apparent reduction of the incidence of atresia as compared to untreated individuals. The amount of atresia quantified in GnRHatreated captive fish was, in fact, similar to the levels measured in wild tuna, but significantly lower to those found in the captive Control group. This finding suggests that the hormonal treatment did not only induce final maturation and ovulation, but by elevating the concentration of the appropriate reproductive hormones it has maintained the viability of the subsequent batches of vitellogenic oocytes, which were the next to undergo final maturation. Such effect of sustained administration of GnRHa been reported in other fishes as well (Mylonas and Zohar, 21a). From the histological analysis of the testes, no differences between GnRHa-treated and Control males emerged in term of reproductive state. Fish at early and late spermatogenesis were equally present both in GnRHa-treated and Control groups. Two spent Control males were found, one in 24 and one in 25 and, in 24, two Controls were in spermiating condition. None of the GnRHa-implanted males was classified as spent, but 5 of them in 25 where spermiating compared to none of he Controls. Altogether, 4 out of 17 Controls (24%) vs 5 among 19 GnRHa-treated fish (26%) were able to release sperm either before or during sampling. The absence of evident effects of GnRHa treatment on testicular histology is in agreement with the absence of statistically significant effects of GnRHa implantation on most sperm characteristics (see next chapter). In conclusion, the histological and stereological analyses reported in the present study indicated that wildcaught female bluefin tuna reared in captivity showed normal oocyte growth, but a low capacity to mature and ovulate spontaneously, while no dramatic differences were observed in response to the GnRHa treatment in the males. Even though spontaneous final maturation and ovulation may not be a likely outcome in bluefin tuna maintained in captivity under the conditions of our study, the results are very encouraging as they demonstrate that (a) males undergo complete spermatogenesis and females complete vitellogenesis, and (b) the achieved maturation stage in females is adequate for their further induction of final maturation and ovulation using hormonal therapies. Furthermore, the results provide hope that under more appropriate conditions (e.g., feeding, stocking density, water quality and longer residence and acclimation to captivity) spontaneous spawning and production of viable embryos may be expected from wild-caught Atlantic bluefin tuna maintained in sea cages, as it has been reported recently for Pacific bluefin tuna (Thunnus orientalis) (Sawada et al., 25) and yellowfin tuna (Thunnus albacares) (Wexler et al., 23). 5.5 EFFECT OF GNRHA TREATMENT ON GAMETE CHARACTERISTICS Introduction 188

189 Male reproduction generally takes place seasonally in temperate fish species. Among a species, the different phases of gametogenesis, that is spermatogenesis, spermiogenesis and spermiation may occur either in a continuum or in different phases separated by resting periods (Billard, 1986). The discontinuity of gamete production is compensated by the immotility of spermatozoa in the genital ducts. The motility is generally triggered by the modification of microenvironment due to ejaculation, for example by the abrupt increase in osmotic pressure in the case of marine fish (Morisawa, 1985). Fish spermatozoa motility duration is related to their ATP content and is very variable among species (Suquet et al., 1994) and may vary along the spermiation period due to sperm ageing (Dreanno et al., 1999). Finally, sperm characteristics may be modified by captivity stress. The use of hormonal treatments has been shown to sustain sperm quality either in case of captivity stress or at the end of reproduction season (Sorbera et al., 1996; Rainis et al., 23). Sperm fertility variations are concomitant to sperm motility modifications (Billard et al., 1986). In recent years, computer assisted sperm analysis (CASA) systems, which were initially developed to examine male fertility in clinical andrology laboratories, have been employed to assess sperm motility of different teleost species (Kime et al., 21; Van Look and Kime, 23). This method provides a fast and objective tool for assessing the quality of the sperm and the possible effects of different environmental conditions. Tuna sperm has been poorly investigated, with only Doi et al. (1982) describing motility and concentration of sperm of wild Mediterranean bluefin tuna. In the present study of REPRODOTT, it was necessary to evaluate a possible stress effect due to captivity on fertility of sperm and to assess a potential effect of hormonal therapy on sperm quality in captive bluefin tuna. The evolution of oocyte diameters in response to hormonal treatment, and the ability of a clearing fixative to determine correctly the stage of maturation of the oocytes was also determined, as part of the original objectives of the study to develop in vitro maturation methods for bluefin tuna oocytes. Due to difficulties in developing such a protocol, however, only the results on the evaluation of maturation are presented in the present report. In the absence of routine production of eggs, sperm fertility was assessed by its motility and concentration characteristics Materials and Methods Ovaries and testes were dissected onboard and brought to the laboratory within one hour. At the lab, three pieces were taken from the dissected ovaries and were dispersed in Ringer s physiological saline and photographed with a computer-controlled digital camera (Leica, Germany) via a dissecting stereoscope. Some oocytes of each sample were treated with a clearing fixative (6:3:1 v/v Ethanol/formaldehyde/acetic acid), which clarifies the cytoplasm and enables evaluation of the maturation stage of the oocytes, by identifying the position of the nucleus and the status cytoplasmic components such as lipid droplets. The physiological stage assumptions after clearing were compared to histological determinations. The diameters of the largest 1% of the oocytes were calculated using image analysis software (Image J, NIH, USA) using the relation Diameter = 2 (Area/PI) assuming that the oocytes were perfect disks. Sperm was drawn by a syringe through an incision of the posterior part of the testes after application of peristaltic pressure along the organ, when needed. Samples of 2 µl of sperm from each male were mixed with 1 ml distilled water and three aliquots were placed on Thoma hemocytometers and were photographed after 5 min decantation. The number of spermatozoa was counted by automatic particle analysis using the image analysis software mentioned above. For motility assessment (only in 25), sperm was first diluted by 1/25 in a isoosmotic Non-Activating Medium (Fauvel et al., 1999). Then it was activated by a dilution of 1/1 in seawater with 1 mg BSA ml -1, placed on a pre-focused Thoma hemocytometer on a compound microscope and viewed/recorded through a 2X objective within 5 s, using a CCD camera (Phillips, The Netherlands). Motility was analysed using a Computer Assisted Sperm Analysis system (Hobson Sperm Tracker, Sheffield, UK) at one minute interval until all movement ceased Results Ovarian staging by the use of the clearing fixative was similar to that of histology (described previously) in all but one case, representing a failure of only a 3%. The use of the clearing solution allowed observing nucleus migration, aspects of the cytoplasm including lipid coalescence and hydration, the latter being characterised by a brownish aspect of vitellus and oocyte burst after contact with the fixative (Fig. 5.19). The increase of mean oocyte diameter after stimulation is mainly due to hydration of competent cells, although a slight significant size increase was observed for previous stages including nucleus migration and lipid coalescence (Fig 5.19). The samples containing oocytes at the stage of hydration show simultaneously vitellogenic and sometimes post vitellogenic oocytes. The mean diameter of the most advanced oocytes was increased significantly by the GnRHa implantation (two-way ANOVA, P <.1), whereas sacrifice/sample time did not have a significant effect (Fig. 5.2). 189

190 A) B) C) D) Fig Ovarian biopsies of captive bluefin tuna after treatment with a clearing fixative. A) control vitellogenic female (ZZ18), B) GnRHa-treated non-vitellogenic female (ZZ11), C and D) GnRHa-treated vitellogenic females showing nuclear migration and lipid droplet coalescence (ZZ 113 & 125). Note that although the diameter of maturing oocytes is increased compared to that of vitellogenic oocytes (arrow), no hydration is observed at this time, hydration being characterised by an oocyte burst at fixation. Immature, primary oocytes, (po), vitellogenic oocytes (vg), maturing oocytes (m). Bar: 5 µm. Control GnRHa ANOVA Parameter P value GnRHa implant.21 Sample time Year ANOVA Parameter P value GnRHa implant.12 Sample time n/a Oocyte diameter (µm) Days after GnRHa implantation Fig. 5.2 Mean (± SEM) sperm density and oocyte diameter of captive-reared bluefin tuna in response to GnRHa implantation during the reproductive seasons of 24 and 25 (n = 4-14, as shown inside the mean bars). In 24, fish were sacrificed 5-6 days after GnRHa implantation, whereas in 25 fish were sacrificed 2-3 days (Cage No 2) or 8 days (Cage No 1) after implantation. There were no statistically significant effects of GnRHa implantation on sperm density, whereas values were significantly lower in 25, compared to 24 (2-way ANOVA, P values on graph). On the contrary, GnRHa implantation had a positive and statistically significant effect on oocyte diameter with the effect of sample time approaching statistical significance (2-way ANOVA, P values on graph). As indicated earlier, the proportion of spermiating males was higher in GnRHa-treated fish than in Controls. However, almost al males, regardless of treatment contained testicular sperm having a concentration ranging from 3 to 4 x1 9 spermatozoa ml -1. Sperm concentration was not affected by the GnRHa implantation (two-way ANOVA, P =.21), but values were significantly lower (P =.1) in 25 compared to 24 (Fig. 5.2). Similarly, there were no significant differences in percentage motile spermatozoa (oneway ANOVA, P =.69), which were 58 ± 9% for Controls and 63 ± 8% for GnRHa-implanted fish (Table 5.7). Finally, the duration of forward motility was long and did not differ significantly between treatments (two-way 19

191 ANOVA, P =.42), and was 9. ± 1.1 min for Controls and 8. ±.7 min for GnRHa-implanted males (Table 5.7). Table 5.7 Mean ± S.E.M of sperm motility characteristics observed de visu by two operators: comparison between Control and GnRHatreated bluefin tuna sperm. The absence of statistical significance is indicated by ns. Control GnRHa implanted Initial motility 58.3 ± 9.4 ns 63.2 ± 8.5 (% motile sperm) n = 6 n = 13 Motility duration 542 ± 67 ns 48 ± 41 (sec) n = 11 n = 15 All the parameters of motility analyzed by CASA showed a very high intra-sample variability. For instance, the most commonly used parameter, the individual curvilinear motility (VCL) representing instant motility, recorded during the first 15 sec ranged from to 25 µm s -1 (mean µm s -1 ) for spermatozoa of the same sample. Then, the variability decreased only due to the slowdown of spermatozoa having higher initial speed. Sperm with motility under 4 µm s -1 were not taken into account by the system (Fig. 5.21). The initial motility at activation concerned from 1 to 9% of the spermatozoa in the GnRHa-treated fish (median 9%) and from 3 to 9% in Controls (median 65%), but there was no significant effect (P =.191) of the treatment (Table 5.7). No significant differences of initial curvilinear motility (VCL) were observed by t- test between Control and treated fish (P=.115). Fig Analysis of motility of individual spermatozoa by Computer Assisted Sperm Analysis (CASA). Each point represents the curvilinear velocity of a spermatozoa recorded for.1 sec. The analysis was performed for the first 15 sec of each minute. The 2-way ANOVA with time and treatment as factors revealed there was a significant effect of GnRHa treatment on sperm overall VCL (i.e., during the whole period of motility, not simply the initial velocity) and average path velocity (VAP) and straight line velocity (VSL) (p<.1) while an effect due to time after activation was detected for VCL and VSL (p<.1) but not for VAP (p=.4) (Fig. 5.22). Plotting the mean VCL (instant speed) and VAP (mean speed on the track) showed systematically higher values along the decay of motility in the sperm of GnRHa-treated fish, compared to that of Controls. Straight line velocity representing the resulting progressive movement speed during tracking time was independent of treatment 191

192 for the first 4 min, whereas thereafter the progression of spermatozoa of GnRHa-treated fish showed a tendency to be higher Discussion The use of the clearing fixative proved to be very efficient to assess instantaneously the maturation status of the gonad in bluefin tuna, as has been the case for other species. This is not to say the histological evaluation should not also follow experimental procedures, but the clearing fixative method can be employed in situations that sacrificing the fish is not possible. In addition, it can be used to decide in situ and after a rapid biopsy, whether a female is a good candidate for hormonal stimulation, since the stage of maturity of the female influences the response to GnRHa treatment (Fauvel and Suquet, 1999; Gardes et al., 2). In the present work, the a priori stimulation of vitellogenic females was successful in most of the cases to trigger the postvitellogenic and maturation processes, driving to a slight increase of oocyte size together with nucleus migration and lipid coalescence. The final increase due to hydration was observed within 2-3 days while new recruitment of competent oocytes was observed at 5-6 and 8 days after implantation. These last observations confirm the reported batch spawning character of bluefin tuna (Medina et al., 22). Bluefin tuna reared in captivity showed general sperm characteristics which were very similar to what was reported for wild fish (Doi et al., 1982). Tuna can be ranked in the fish with high sperm concentration compared to other seawater fish, while sperm motility duration is intermediate between some extreme cases such as turbot (Scopthalmus maximus)(long motility) and European seabass (short motility) (Suquet et al., 1994). Due to the high swimming speed of this pelagic fish and the possible loss of a large amount of ejaculated sperm during mating, high concentration and long motility may be a strategy to maintain reproduction success. Whereas no effect of GnRHa treatment was revealed by the simple observation of sperm motility, the use of CASA allowed analyzing more precisely the effect of hormone application. In fact, instant and average speed and progressive VSL, as well as motility duration were increased in GnRHa-treated fish, while no significant difference was observed for concentration. Such a result is original since, generally, the effect of hormones are in relation to concentration and volume of available milt (Sorbera et al., 1996), without major effectw on sperm motility. The results of the present study suggest that the developed GnRHa-implantation therapy may increase bluefin tuna male fertility in cage conditions, through improvement of spermiation and sperm motility characteristics. 192

193 VCL (µm s -1 ) control GnRHa n= time post activation (min) 1 VAP (µm s -1 ) 8 6 n= time post activation (min) 4 VSL(µm s -1 ) 3 2 n= time post activation (min) Fig Comparison between control and GnRHa-treated male sperm motility by Computer Assisted Sperm Analysis (CASA). The figure shows the decay of the three mean population parameters of sperm motility with time for 3 GnRHa-treated and 3 Control fish. The mean population motility parameters of each individual are established using a maximum of 1 sperm tracks per minute. 193

194 SYNERGY OF THE CHAPTERS AND CONCLUSIONS As it was pointed out at the Introduction, the four main objectives of the project were: 1) To improve the knowledge of the reproductive biology of the species in captivity and compare this with wild populations, in order to develop an aquaculture farming technology for the bluefin tuna (BFT) 2) To assess the capability of BFT broodstock to mature and spawn in captivity 3) To obtain viable eggs from BFT breeders and bring them to hatching 4) To develop handling techniques for routine operations in BFT aquaculture and research The project activities were designed to achieve these objectives by splitting the work into several work packages according with different topics of research and which its output could be gathered to contribute in reaching the goals. These section is willing to summarise a provide a clear idea about how the different results obtained in the REPRODOTT project have facilitated of attaining the objectives, IMPROVED THE KNOWLEDGE OF THE BFT REPRODUCTIVE BIOLOGY Along the three years of the project it was carried out analysis on samples from about 6 fish captured from the wild by different fishing methods (purse seine, long-line and traps), at the Eastern Atlantic Ocean and in different Mediterranean areas: Balearic Islands, North Ionian Sea, South Tyrrhenian Sea, Sardinia Chanel, Malta offshore waters, North Aegean Sea and Levantine Sea. Besides, another 2 captive BFT were sampled from the experimental facilities and some BFT farms located in Turkey. All the teams used the same methodology with the sample, so that the results would be comparable. With data concerning the size and weight of the specimens, a morphometrical study was carried out. Furthermore, with samples taken from the blood, brain, pituitary, gonads, and sexual products a detailed analysis was carried out by the different partners responsible for the project. As results of the different analyses and studies made on these samples: a) An extensive database has been generated from the BFT populations studied composed by biometric indices, histological features of the gonads, levels of reproductive hormones, characteristics of gametes and stress parameters. b) These data have defined a complete profile of the maturation process of the BFT in the Mediterranean, information that was inexistent up to now us, and provided the necessary basis to characterize the reproductive biology of BFT and their performance in wild and captive conditions. c) New tools and procedures, specific methods of analysis of the hormones related with reproduction (GnRHs, GTHs, vitellogenin), have been developed that may facilitate further in depth studies on BFT reproduction d) New basic scientific information have been obtained on the brain-pituitary-gonad axis related with the reproduction such as the sequencing and characterization of brain reproductive hormones and pituitary hormones; the seasonal profiles of brain pituitary and gonad reproductive hormones; the BFT gametes characteristics, etc. Through histological and hormonal analyses, it was obtained some consistent results which may have a clear interest for the management and conservation of BFT stocks. Particularly: The identification of a new spawning ground of BFT in the Levantine Sea and the confirmation of the existence of three others around the Balearic Islands, South Tyrrhenian Sea and Malta offshore waters. The finding that spawning season of BFT occurs about one month earlier in the Levantine Sea than in the Central and Western Mediterranean. The fact that male BFT specimens attain in the Mediterranean a larger size, grow faster and have a major longevity than females. The fact that Eastern Atlantic BFT female has a size and age at first sexual maturity markedly lower than the Western Atlantic stock. The proof that the assessment of wild BFT reproductive status is strongly influenced by the fishing gears employed for sampling, being long line fishery not suitable for estimating certain reproductive parameters (e.g. fecundity and spawning frequency). 194

195 So, it can be concluded that the project has improved notably our knowledge of the Reproductive Biology of Atlantic bluefin tuna (BFT), Thunnus thynnus thynnus, both in captive and under wild conditions, throughout the Mediterranean area. ASSESSED THE CAPABILITY OF BFT TO MATURE AND SPAWN IN CAPTIVITY Several behavioural studies, histological and hormonal analyses carried out in two consecutive years on the reproductive status of captive breeders, kept in floating cage for over two years, have produced consistent results, which clearly indicate that BFT are able to mature and spawn in a natural way after 2 or 3 years of captivity conditions. With regard to the comparative study between wild and captive tuna; although some of the specimens in captivity reached maturity and spawned eggs, it seemed that it was necessary to induce final maturation through the use of hormonal treatment. Thus, captive fish showed similar levels of brain and pituitary hormones than wild BFT and dysfunctions seemed to be more on the vitellogenesis process and therefore, related on the gonad hormones (sex steroids). This indication is consistent with the results on the morphometry (relative size of the gonads respect of fish body weight) and histological studies, and hormonal measurements. ASSESSED THE FEASIBILITY TO OBTAIN VIABLE EGGS FROM BFT BREEDERS AND BRING THEM TO HATCHING The development of a suitable hormonal delivery system and its implantation procedure, and the lessons acquired after two years of attempts to control the BFT reproductions have resulted in the successful production of eggs, which were then brought to hatching. Thus, on July 7th 25 a number of viable eggs were obtained, which after being artificially fertilised with sperm from one male led to the achievement of ABT larvae in a controlled way and for the first time ever at worldwide level. Additionally, it has been developed a very useful tool for fish reproductive studies, a sensitive quantitative and qualitative determination assays for the hormonal inducer in the fish blood. Furthermore, the production of a suitable hormonal implant for inducing spawning of BFT might be also useful to be employed in other large fish species, like tunas. Sacrificing hormonally treated for collecting ovulated eggs from the ovaries to undertake in vitro fertilization, it is obviously a method that can not be used for the routine acquisition of fertilized eggs from captive broodstock, as it requires the destruction of the animal. However, as long as the industry of bluefin tuna aquaculture is based in the collection of mature fish from the wild, it is possible to maintain a few fish in a smaller cage and obtain eggs for the development of larval rearing methods, by administering GnRHa implants during the reproductive season, and then obtaining the eggs and sperm after sacrificing the fish at a predetermined time. In this way, further progress can be made in the various stages of rearing of bluefin tuna in captivity, concurrently with our efforts to develop spontaneously spawning, captive broodstock. DEVELOPED HANDLING TECHNIQUES FOR ROUTINE OPERATIONS IN BFT AQUACULTURE Some progresses were made in the setting up of anaesthesia procedures, transportation systems, the design of egg collectors for rearing cages, employment of non-invasive techniques for sex and maturity assessment, and for conducting behavioural studies through echo-sounding and data logger tagging. Future action on the anesthesia approach should be focused on attempting new trials by employment of other application procedures and testing more powerful and quick effect drugs. Regarding egg collection devices, it should be tested the one developed lately in cages with much more broodstock specimens to check with more accuracy its effectiveness. On the non-invasive techniques approach for sex and maturity determination, further actions should be concentrated on improving the design of the dot blot and making is available for commercial use either within the ERA or also widening the remit for an international market. It is suggested that further the developments of the attachment head and tag could be carried out and that in future programmes involving brood stock data-loggers could be employed for monitoring environmental changes. Moreover, the technical improvement of echo-sounding approach should be tackled. GENERAL CONCLUSION The main achievement of the REPRODOTT project was the obtaining of fertilized eggs and hatched larvae from BFT broodstock alter hormonal treatment to induce final maturation and spawning. This has given use the final confirmation that BFT in captivity can mature and spawn fertile gametes. Moreover, this result is a 195

196 major breakthrough, since it s the first time ever, at worldwide level, that the reproduction of Atlantic BFT has been achieved in a controlled way. Although to further progress into the full domestication of BFT, all the researchers participating in the REPRODOTT project recognise that the need to improve the techniques developed and employed implies the availability of specifically designed research infrastructure, which until now has been inexistent in Europe; although not so in other parts of the world (Panama, Australia, Japan, Indonesia). An infrastructure of these characteristics would allow some of the difficulties found in the project to be resolved, such as the collection of eggs, and progresses in the development of adequate handling techniques for the tuna, which both result fundamental for a self-sustained BFT aquaculture and research. 196

197 DISSEMINATION OF RESULTS EXPLOITATION AND DISSEMINATION OF RESULTS It has been always the intent of the project consortium to strive for communication and dissemination at the widest and highest level possible, to ensure that international cooperation and interaction could be achieved. Thus, the dissemination of the objectives and results of the project have been made at different levels targeting different sections of publics: Dissemination towards the scientific community Scientific Publications The main results obtained in REPRODOTT have been disseminated to the scientific community and will continue through publication in peer-reviewed international journals (with explicit reference to the supporting role of EU by funding this research). Results from the study were evaluated and organised for presentation and publication during the annual project meetings, under the supervision of the responsible scientists from the involved partners. To date, the following scientific papers have been either published or submitted to perreview international journals: Rosenfeld, H., Klenke, U. Gordin, H. And Zohar, Y. 23. Preliminary study of the key hormones regulating reproduction in the bluefin tuna (BFT): the brain gonadotropin-releasing hormones (GnRHs) and the pituitary gonadotropins, FSH and LH. Cah. Options. Medit. 6: Abascal, F. J., Megina, C. and Medina, A. 24. Testicular development in migrant and spawning bluefin tuna, Thunnus thynnus (L.), from the eastern Atlantic and Mediterranean. Fishery Bulletin 12(3): C. Fauvel & M.Suquet, 24. La domestication des Poissons: Le thon rouge INRA Prod. Anim. 17, Karakulak S., Oray I., Corriero A., Aprea A., Spedicato D., Zubani D., Santamaria N., De Metrio G. 24 First information the reproductive biology of the bluefin tuna (Thunnus thynnus) in the Eastern Mediterranean. Col. Vol. Sci. Pap. ICCAT, 56(3): Karakulak S., Oray I., Corriero A., Deflorio M., Santamaria N., Desantis S., and De Metrio G. 24 Evidence of a spawning area for the bluefin tuna (Thunnus thynnus L.) in the Eastern Mediterranean. Journal of Applied Ichthyology 2: Corriero A., Karakulak S., Santamaria N., Deflorio M., Spedicato D., Addis P., Desantis S., Cirillo F., Fenech-Farrugia A., Vassallo-Agius R., de la Serna J.M., Oray Y., Cau A., De Metrio G. 25.Size and age at sexual maturity of female bluefin tuna (Thunnus thynnus L. 1758) from the Mediterranean Sea. Journal of Applied Ichthyology, 21: Garcia, A Estudio de factibilidad para el control de la reproducción del atún rojo en cautividad, REPRODOTT (Q5RS ). El Cimarrón del Atlántico Norte y Mediterráneo. J.L. Cort (Ed.). Instituto Español de Oceanografía. Santander 25. pp.: Abascal F. J. and A. Medina. 25. Oogenesis ultrastructure in the bluefin tuna, Thunnus thynnus. Journal of Morphology 264: De la Serna, J.M., A. Garcia, A. Garcia, J.M. Ortiz de Urbina y D. Macias. 26. Infrome sobre las actividades de investigación sobre atún rojo desarrolladas por el instituto Español de Oceanografía en el Mediterraneo dentro del Programa BYP de ICCAT. Col. Vol. Sci. Pap. ICCAT, 59 (3): De la Gandara, F., J. Miquel, M. Iglesias, A. Belmonte, E. Ayora and A. Garcia-Gomez. 26. Using an echosounder system to study the vertical movements of captive bluefin tuna (Thunnus thynnus) in floating cages. European Aquaculture Society Special Publication, 35: Garcia A., M.V. Díaz, F. De la Gándara, J.M. De la Serna, A. Belmonte, E. Ayora, H. Gordin, C. Fauvel, A. Medina, C. R. Bridges, R. Vasallo-Agius, C. Mylonas y G. Demetrio. 25. Posibilidades de reproducción del atún rojo, Thunnus thynnus, en cautividad. Actas X Congreso Nacional de Acuicultura, Tomo II: Medina, A., F.J. Abascal, L. Aragon, G. Mourente, G. Arnada, T. Galaz, A. Belmonte, J.M. De la Serna y S. Garcia. 27. Influence of sampling gear in assessment of reproductive parameters fro bluefin tuna in the western Mediterranean. Marine Ecology Progress Series, 337: Mylonas, C.C., C.R. Bridges, H. Gordin, A. Belmonte Ríos, A., García, F. De la Gándara, C. Fauvel, M. Suquet, A., Medina, M. Papadaki, G. Heinisch, G. De Metrio, A. Corriero, R. Vassallo-Agius, J.M. Guzmán, E. Mañanos and Y. Zohar. 28. Preparation and administration of gonadotropin-releasing hormone agonist (GnRHa) implants for the artificial control of reproductive maturation in captive-reared Atlantic bluefin tuna (Thunnus thynnus thynnus). Reviews in Fisheries Science (in press). 197

198 Mylonas, C.C., M. Papadaki, C.R. Bridges, H. Gordin, G. Heinisch, A. Belmonte-Ríos, A. García, F. de la Gandara, C. Fauvel, M. Suquet, G. De Metrio, A. Medina, R. Vassallo-Agius, J.M. Guzman, E. Mañanos and Y. Zohar. 26. Artificial control of reproductive maturation in captive-reared bluefin tuna (Thunnus thynnus): I. Preparation and administration of gonadotropin-releasing hormone agonist (GnRHa)-loaded implants and induction of ovulation. Aquaculture (submitted) The followings are in preparation: Mylonas, C.C., M. Papadaki, C.R. Bridges, H. Rosenfeld, G. Heinisch, Y. Zohar, F.J. Abascal and A. Corriero. Induction of maturation in captive-reared bluefin tuna (Thunnus thynnus): II. native GnRHs, luteinizing hormone, sex steroid hormones and gonadal development after implantation with GnRHa. Aquaculture (in preparation) Corriero, A., A. Medina, C.C. Mylonas, F.J. Abascal, M. Deflorio, L. Aragon, C.R. Bridges, N. Santamaría, G. Heinisch, R. Vasallo-Agius, A. Belmonte, C. Fauvel, A. Garcia, H. Gordin y G. De Metrio. Histological study of the effects of treatments with gonadotropin-releasing hormone agonist (GnRHa) on reproductive maturation in captive-reared Atlantic bluefin tuna (Thunnus thynnus thynnus L.) Contributions to Scientific Conferences and Symposiums General presentations of the REPRODOTT project and selected results (lectures, posters) were made by the project participants at relevant national and international conferences and/or workshops. It must be pointed out that some REPRODOTT partners were invited by Japanese and Australian institutions to give lectures about main REPRODOTT achievements at international conferences organised by them. These have been the conferences where the REPRODOTT results have been disseminated during the project: International Conferences - 5 Guest Lectures at International Conference on Advances in Biotechnologies Applied to the Reproduction of Captive Fish: New Experiences and Related Problems. Held at University of Bari, Faculty of Veterinary Medicine, Department of Animal Health and Well-being in Valenzano, Bari (Italy), March 24: C. Fauvel & M. Suquet. Gamete management in Aquaculture. H. Gordin. 24. New experiences and related problems to the domestication of Thunnus thynnus, the blue fin tuna. H. Rosenfeld. The central regulators of the reproductive axis: From genes to proteins. C. Mylonas. Reproductive dysfunctions of cultured fish and hormonal manipulations for the induction of spawning C.R. Bridges. Sex determination and Tagging of Individual Broodstock in Captivity - Methods and Problems. - 1 contribution to the European Aquaculture Society Meeting: Aquaculture Europe 24: Biotechnologies for quality. Held in Barcleona (Spain), 2-23 October 24: C.R. Bridges. Broodstock Identification: Sex, Sexual Maturity and Individuality in Bluefin Tuna - 3 Guest Lectures at 21st century COE International Symposium "Stock Enhancement and Aquaculture Technology. Held at Kinki University, Osaka (Japan) 1-11 November 24: A. Garcia, A. Belmonte, E. Ayora, H. Gordin, C. Fauvel, A. Medina, C. Bridges, R. Vasallo-Agius, C. Mylonas & G. De Metrio. Bluefin Tuna in the Mediterranean: Reproduction In Captivity C. Fauvel, M. Suquet, F. de la Gandara, A. Medina, F. Abascal & C.Mylonas, 24. Gamete biology: Perspectives for Bluefin Tuna Aquaculture. G. De Metrio, G.P. Arnold, J.M. de la Serna, J.L. Cort, B.A. Block, P. Megalofonou, M. Lutcavage, I. Oray, M. Deflorio. Observations on migratory behaviour of Atlantic bluefin tuna (Thunnus thynnus L.) tagged with pop-up satellite tags in the Mediterranean Sea. - 3 Contributions to the International Symposium World Aquaculture Society 25: international peace and development through Aquaculture. Held in Bali (Indonesia) Mayo 25: Díaz M.V., A. Garcia, C. Bridges, A. Belmonte, A. Medina, C. Mylonas y H. Gordin. Development of handling procedures for captive bluefin tuna Thunnus thynnus. Garcia, A., A. Belmonte, E. Ayora, H. Gordin, C. Fauvel, A. Medina, C. Bridges, R.Vasallo-Agius, C. Mylonas y G. Demetrio. A feasibility study for the domestication and reproduction of Bluefin Tuna Thunnus thynnus in captivity: the REPRODOTT project. 198

199 Mylonas, C.C., Belmonte, A., Garcia, A., Bridges, C., Rosenfeld, H., Medina, A., Corriero, A., Demetrio, G., Fauvel, C., Vassallo-Agius, R., Zohar, Y. and H. Gordin. Spawning induction of captive-reared bluefin tuna (Thunnus thynnus) using GnRHa implants. - 1 Contribution to the International Symposium Aquaculture Europe 25. Lessons from the past to optimize the future. Held in Trondheim (Norway), August 25: De la Gándara F., J. Miquel, M. Iglesias, A. Belmonte, E. Ayora and A. García-Gómez. Using an echosounder system to study the vertical movements of captive bluefin tuna (Thunnus thynnus) in floating cages. - 6 Contributions to the International Symposium World Aquaculture 26. Held in Florence (Italy), 9-13 May, 26: Corriero, A. De Metrio, G., Abascal, F. and Medina, A. Effects of GnRHa on the gonads of captive-reared Atlantic Bluefin tuna Thunnus thynnus thynnus. Fauvel, C. Abascal Crespo, F. J., de la Gandara, F.,Suquet, M. and Cosson, J. Bluefin tuna Thunnus thunnus thynnus gamete maturation after hormonal induction. Heinisch, G., Corriero, A., Medina, A. Abascal, F. J., De la Serna, J. M., Vassallo Aguis, R., Belmonte, A., Garcia, A., De la Gandara, F., Fauvel, C., Bridges, C. R. Mylonas, C. C., Oray, I., De Metrio, G., Rosenfeld H. and Gordin, Hillel. Spawning behaviour of bluefin tuna, Thunnus thynnus, across the Mediterranean Sea. Gordin, H. Where should bluefin tuna R&D go? Mylonas, C.C., Belmonte, A., Garcia, A., Bridges, C., Rosenfeld, H., Medina, A., Corriero, A., Demetrio, G., Fauvel, C., Vassallo-Agius, R., Zohar, Y. and H. Gordin. Induction of ovulation and production of viable larvae of bluefin tuna (Thunnus thynnus) using GnRHa implants. Rosenfeld, H., Klenke, U., Gliksman, D., Meiri, I., Heinisch, G., Corriero, A., Vassallo Aguis, R., Belmonte, A., Garcia, A., Medina, A., De Metrio, G., Mylonas, C. C., Zohar, Y., Gordin, H. and Bridges C. R. Domestication of bluefin tuna (Thunnus thynnus) - a vision or reality?! - 3 Guest lectures to Skretting Australasian Aquaculture Conference held in Adelaida (South Australia) from 27 to 3 August 26: Gordin, H. and C.C. Mylonas Fattening blue fin tuna in the Mediterranean Sea Mylonas, C.C. et al. Induction of ovulation and production of viable larvae of bluefin tuna Thunnus thynnus using GnRHa. Rosenfeld, H., G. Heinisch, I. Meiri, D. Gliksman, A. Corriero, A. Medina, F.J. Abascal, J.M. de la Serna, R. Vassallo Aguis, A. Belmonte, A. Garcia, F. de la Gandara, C. Fauvel, C.C. Mylonas, I. Oray, G. De Metrio, C.R. Bridges, Y. Zohar, and H. Gordin. Spawning activity of bluefin tuna (Thunnus thynnus) in the Mediterranean Sea: wild vs captive broodstocks. - 2 Guest Lectures at 21st century COE International Symposium "Ecology and Aquaculture of Bluefin Tuna. Held at Amami, Oshima (Japan), November 26: Garcia, A. General overview of the EU tuna research problems and issues. Mylonas, C.C., C. Bridges, A. Belmonte, A. Garcia, C. Fauvel, H. Rosenfeld, A. Medina, A. Corriero, G. De Metrio, R. Vasallo-Agius and Y. Zohar. Artificial induction of maturation and spawning of Atlantic Bluefin tuna, Thunnus thynnus. - 2 Contributions to the 8 th International Marine Biotechnology Conference. Held in Eilat (Israel), March, 27: H. Rosenfeld, G. Heinisch, I. Meiri, D. Gliksman, A. Corriero, A. Medina, F.J. Abascal, J.M. de la Serna, R.V. Agius, A. Belmonte, A. Garcia, F. de la Gandara, C. Fauvel, C. Mylonas, I. Oray, G. De Metrio, C.R. Bridges, Y. Zohar and H. Gordin. Spawning activity of bluefin tuna in the Mediterranean sea: Wild vs. Captive broodstocks H. Gordin. Fattening blue fin tuna in the mediterranean sea problems and possible solutions - 1 Guest lecture at European Tuna Conference, held in Brussels on April 23 rd 27: García, A. Tuna hatching - 2 Contributions to the 8 th International Symposium on Reproductive Physiology of Fish. Held in St Malo (France), 3-8 June 27: C. Mylonas C.Induction of spermiation, ovulation and spawning in Atlantic bluefin tuna (Thunnus thynnus) using GnRHa delivery systems. 199

200 H. Rosenfeld. Reproductive activity of bluefin tuna (Thunnus thynnus) in the Mediterranean Sea: wild vs. captive broodstocks. National Conferences - Contributions to the Seminar: Acuicultura I: Biología Marina. Reproducción y Desarrollo. Aulas del Mar. Universidad Internacional del Mar. Universidad de Murcia. Held in Cartagena (Spain), September, 23: Medina, A. Perspectivas del cultivo del atún rojo, Thunnus thynnus: Reproducción en cautividad - Contributions to the Seminar: Acuicultura I: Biología Marina. Reproducción y Desarrollo. Aulas del Mar. Universidad Internacional del Mar. Universidad de Murcia. Held in Cartagena (Spain), September, 24: Medina, A. Perspectivas del cultivo del atún rojo, Thunnus thynnus: Reproducción en cautividad - Contributions to the X Congreso Nacional de Acuicultura. Held in Gandia (Spain), October 25 Garcia A., M.V. Díaz, F. De la Gándara, J.M. De la Serna, A. Belmonte, E. Ayora, H. Gordin, C. Fauvel, A. Medina, C. R. Bridges, R. Vasallo-Agius, C. Mylonas y G. Demetrio. Posibilidades de reproducción del atún rojo, Thunnus thynnus, en cautividad. Libro de Resúmenes. Tomo II: Contributions to the Seminar: Acuicultura I: Biología Marina. Reproducción y Desarrollo. Aulas del Mar. Universidad Internacional del Mar. Universidad de Murcia. Held in Cartagena (Spain), September, 25: Medina, A. Perspectivas del cultivo de túnidos: reproducción en cautividad. - Contributions to the 12th Pan-Hellenic Symposium of Ichthyologists. Held in Drama (Greece), October 25: Mylonas, C.C., Belmonte, A., Garcia, A., Bridges, C., Rosenfeld, H., Medina, A., Corriero, A., Demetrio, G., Fauvel, C., Vassallo-Agius, R., Zohar, Y. and H. Gordin. Control of reproductive maturation of bluefin tuna (Thunnus thynnus) using GnRHa implants and production of larvae using in vitro fertilization. - Contributions to the 8 th Panhellenic Symposium of Oceanography and Fisheries. Held in Thessaloniki (Greece), 4-7 June 26: Mylonas, C.C. and M. Papadaki. Control of reproductive maturation of bluefin tuna (Thunnus thynnus, Linnaeus) using GnRHa implants, and production of larvae with in vitro fertilization. - Contributions to the Workshop on Biochemistry and Genetic of large pelagic fish. Held in Genoa (Italy), 29 October 24 De Metrio, G. Recent observations on the reproduction and movements of bluefin tuna in the Mediterranean. REPRO-DOTT WWW home page A World Wide Web site was established for the REPRODOTT. The REPRODOTT at the beginning of the project and is actually accessible at the following address: The website has evolved over the duration of the project reflecting its progress. The pages describe the project and progress to-date, and are aimed at both scientists and a non-scientific audience. Via the website, tuna scientists, stakeholders and interested parties have open access to the deliverables, photo gallery, and parts of the progress reports which are not of restricted nature due to intellectual property protection and scientific publication concerns. There is restricted access to some pages with some confidential information and which allows partners to exchange data and report progress. The website will be available beyond the life of the project allowing project outputs to be assessed and material to be downloaded over a longer time period and to reach a wider audience. Invitation to annual project meetings The Project Consortium decided to invite scientists and professionals involved with the biology and aquaculture of the bluefin tuna in Europe and the rest of the world as well as international management bodies. Our objective was to (a) disseminate the findings of our research to relevant people as soon as possible, and (b) to gain from the experience of these participants in order to encourage progress in the REPRODOTT work. There were guests participants from public research institutions and private industry from Mediterranean countries (France, Italy, Turkey, Netherlands, Cyprus, and Malta). As well as representatives from Australia 2

201 (SARDI Aquatic Sciences), and International organisations such as EC-DG Fisheries, IATTC, FAO attending the three REPRODOTT annual meeting held in France, Greece and Malta. Dissemination towards European Commission Apart form the annual interim report submitted, the invitation and participation of scientific officers from the DG Fish of European Commission to the annual steering committees have ensured that it would be regularly informed of the progress of the project as well as main outputs. Moreover, a Lunch conference on the results and achievements of the REPRODOTT project was given in December 25 by the coordinator and a few more partners at DG Fish facilities in Brussels. Furthermore, a press release on first successful achievement of in vitro fertilisation of Bluefin tuna eggs was published in the CORDIS wire releases web page in July 25. Also, the project partners answered a survey conducted within the framework of IMPACT FISH (EC project FP6-23-SSP ), a quantitative impact assessment of project outputs of FP4 and FP5 Research Programmes on Fisheries, Aquaculture and Seafood Processing Research Area and the Fisheries Industry. In this survey, a summary of the scientific and technical publications, training and mobility of personnel, new processes and methods developed, and disseminations as results of the project up to date was givem. The results were published Dissemination towards ICCAT Several REPRODOTT scientists have participated into the ad hoc GFCM/ICCAT Working Group on sustainable Tuna Farming/Fattening Practices in the Mediterranean. This working group was set up following a 22 decision by the General Fisheries Commission for the Mediterranean (GFCM), which in view of the expansion of the bluefin tuna farming in the Mediterranean, decided that practical guidelines to ensure the sustainability of this activity were required. The group was composed by 19 experts from 1 Mediterranean countries, Japan, the European Commission, and representatives from the Secretariats of GFCM and ICCAT. It helds two meetings in 23: May at Rome (Italy) and December at Izmir (Turkey), and then a final meeting in Rome (Italy) from 16 to 18 March 25. As results, a snapshot of the current situation of Mediterranean bluefin tuna capture fisheries, farming and marketing/trade was reported. This helped to the drafting and adoption of the Guidelines on sustainable Tuna Farming Practices in the Mediterranean (see FAO Fisheries Report, 779. Rome, FAO p.) Apart from the contribution to the working group task, an overview of the REPRODOTT project was provided to the meeting participants. Dissemination towards fishermen, NGOs and civil society DVD documentary A 33 minutes DVD film which contains a summary of activities and achievements carried out during the REPRODOTT project have been released in 27 to serve for the dissemination among stakeholders and the general media. It has been produced by Accion Visual, a communication company, in two versions (English and Spanish). Popular articles and press, radio and television interviews Some press articles have been published at national and international journals by interview with either the project coordinator or representatives from the different project partners. In that sense, it must be pointed out an article published in the 23 May number of Fish Farming International, the leading international publication of the aquaculture industry. Another article was published at the same journal in August 25 describing the major achievement during the third year of the project, which included the successful induction of maturation and spawning in captive bluefin tuna, and the production of viable larvae using both spontaneous spawning and in vitro fertilization 21

202 Several articles have also been published on the same subjects at the magazine Fisheries News, a Greek magazine which is well-read by scientists, regulators and industry personnel in the field of fisheries and aquaculture. Other outputs, such as popular articles to other fish farming magazines and established web-sites were generated more frequently and disseminated at the international and national level by all partners. In particular, a press release informing about the success on controlling the reproduction and to obtain bluefin tuna larvae was made by the project consortium which was published in many national and international magazines and specialised web sites. Besides that, some interview on radio specialized programs to explain the aim and activities of the project were made. Together with some cover articles or news notes on local or national newspapers and interviews on regional televisions. EXPLOITATION OF RESULTS The REPRODOTT project has generated several tool and assays that could be employed on further researches. For instance, the specific ELISA for BFT GnRH and GTHs can now be used on BFT reproductive biology studies. It has been developed an ELISA for measuring blood plasma concentration of GnRHa that could be used in BFT and other marine fish to asses the effectiveness of hormonal induction experiments of maturation and spawning. Moreover, a new technology (modified Dot-Blot for vitellogenin and zona radiata proteins) has been set up for identification of male and female BFT. Another important achievement has been the production of a slow-release hormonal (GnRHa) delivery system (implant) for BFT and other large pelagic fish. Already these technical results have been commercialized and are being taken up by the Australian Tuna fattening industry. Similar consultancy work is planned for national EU programmes in Italy in