Abstract Atlantic bluefin tuna (Thunnus thynnus) are currently managed as two separate stocks, one breeding in the Mediterranean Sea and the other in the Gulf of Mexico. The fidelity of the two stocks to their respective breeding ground is currently in question. In the Gulf of Mexico, bluefin arrive in a sexually inactive state and in some time period ranging from weeks to months, develop into spawners. This western stock is reported to mature at about 190-205 cm in length and about 8-10 years of age. Currently, bluefin tuna are being tagged with pop-up satellite and archival tags, to discern their migratory movements and philopatry to a spawning ground. My objective was to examine bluefin in the areas they are currently being tagged, to discern the state of sexual maturity of fish in the different regions. Specifically, I was interested in identifying 1) what is the minimum size of first sexual activity and 2) if spawners were restricted only to the Gulf of Mexico. Knowing which fish are mature will help to interpret the movement and behaviors revealed by the tag data. Since scombrid fishes are closely related, other species provided a valuable comparison. To histologically identify the stages of wild caught bluefin tuna I compared the gonadal tissues to those of yellowfin tuna (Thunnus albacares) and Eastern Pacific bonito (Sarda chiliensis) held in captivity at the Tuna Research and Conservation Center. Based on a classification system from previous studies, I assigned each sample a stage of maturity. In addition, for captive fish, I collected data on gonad size at mortality, which provided an index of gonad development for corroboration with histological examinations. Ifound that female yellowfin tuna do not become sexually active in water temperatures below 20°C, but male yellowfin do, with no apparent variation in size to first sexual maturity from wild caught yellowfin tuna. Female yellowfin at 25°C showed slightly more advanced oocytes than females at 20°C. Both sexes of bonito become sexually active in captivity and spawn in 20°C water. Together, histological observations of the captive populations reveal all the stages of gonadal development needed to classify scombrid gonad samples. Based on comparative histological analyses, I concluded that all bluefin gonadal tissues (n-111) from New England and North Carolina were sexually inactive and therefore not spawning. All fish from the Gulf of Mexico and the Bahamas (n=5) demonstrated asynchronous ovaries with many oocytes yolked and in maturation stages just prior to spawning. The lack of fully hydrated oocytes, indicative of a currently spawning fish is probably due to the small sample size and time of capture. The two Bahama fish represent the first histological observations of sexually active bluefin tuna outside the Gulf of Mexico. Understanding the reproductive biology and the size at first spawning is a crucial step for developing future management strategies. Introduction Tuna are an epipelagic group of fish (family Scombridae) that inhabit every major ocean and support commercial and sport fisheries around the world. They are known for their unique physiological traits which include a hydrodynamic shape, high degree of endothermy, as well as counter-current heat exchangers (Carey 1969). In some species such as the Atlantic bluefin tuna, these characteristics allow them to migrate thousands of miles from warm temperate spawning grounds to food-rich colder waters. Despite the intriguing physiology and economic importance of these species, mysteries still remain about their biology and behavior, especially the reproductive dynamics which are crucial for assessing stock structure and managing the respective fisheries. Although scombrid fishes all have extremely similar stages of sexual development, the factors that govern this development are different for each species. The species considered in this study include the Atlantic bluefin tuna (Thunnus thynnus thynnus) the yellowfin tuna (Thunnus albacares) and the Easter Pacific bonito (Sarda chiliensis). Atlantic bluefin tuna are most reknown for their large sizes (up to 600 kg) and for endothermy which allows them to spend extensive amounts of time in colder waters. Mainly due to overfishing, there has been a drastic decline in the bluefin stocks (-10% of 1975 levels). They are currently managed as two separate stocks. The eastern stock spawns in the Mediterranean at approximately 5-6 years of age or 150 cm (Rodriguez- Roda 1967) and the western stock spawns in the Gulf of Mexico at approximately 8-10 years of age or 190-205 cm (NRC report 1994, Baglin 1982). Recently, alternative spawning sites have been proposed but very little evidence exists for any kind of sexual activity outside of the Gulf of Mexico and the Florida Straits. Western Atlantic bluefin are seasonally distributed along the American coast from 60° N to 20° N and spawning season in the Gulf lasts from April through June (Rivas 1954, Baglin 1982, McGowan and Richards 1989). Based on tag returns, the mixing between stocks is believed to be approximately 5% (Calaprice 1986). A preliminary study on the degree of spawning site fidelity (Nemerson et al. 1998) has shown no significant mixing between the spawning grounds. Because the literature regarding these questions is sparse, studies utilizing new technologies are continuing to reveal new behaviors. The amount of spawning site fidelity and inter-stock mixing is currently being determined by tagging projects using pop-up satellite archival tags (PSAT) and internal archival tags (Block et al. 1998). To be most efficient in deploying the tags, these projects target only sexually mature fish that will be going to the spawning grounds within the year. Since the literature regarding the age at which bluefin first become sexually mature contains a great deal of variation, I felt that further investigation was necessary. Baglin (1982) presents the most extensive study to date of the reproductive biology of the western stock. He saw very little evidence of sexual maturity in age 1-7 year old bluefin in the Mid-Atlanic Bight. The only area where sexually active bluefin were found was in the Gulf of Mexico and Florida Straits. This study was later criticized by Clay (1990) due to its relatively low sample sizes in different regions and seasons. Data collected by National Marine Fisheries Service (NMFS) observers in the Gulf of Mexico show that bluefin are caught only during spawning season and that only fish greater than 135 kg (190 cm) migrate there. Based on this information, the current theory is that all bluefin arriving in the Gulf are there to spawn (Nemerson et al. 1998). Rivas (1954) gives evidence of the Florida Straits also being a spawning site, based on behaviors, gross morphology of gonads, egg diameters, and larval surveys, but provides no histological evidence. Bluefin taken in this region range from 135 to 320 kg (200-270 cm). Yellowfin tuna are known to feed and spawn in tropical and subtropical regions worldwide. They are targeted by a purse seine fishery in the Eastern Tropical Pacific (EPO) and are harvested at an average of 216,300 metric tons per year. Schaefer (1996. 1998) has studied the yellowfin in the EPO from 30° N to 15° S with the majority of yellowfin sampled between the equator and 20° N. In the 1998 Inter-American Tropical Tuna Commission (IATTC) report, Schaefer describes a significant correlation between Sea Surface Temperature (SST) and proportion of reproductively active fish in all four of the areas sampled. Fish in spawning condition were encountered in a temperature range of 21.5° to 30.0°C, however, 85.3% of spawning occurred between 26.0° to 30.0° C. Schaefer also determined that 50% of male yellowfin were sexually mature at 69.0 cm and the smallest was less than 50 cm. Females were larger before becoming sexually mature, the smallest being 59 cm and 50% were mature at 92.1 cm. Yellowfin in the EPO spawn at night between 2200 h and 0600 h. As shown in this study, yellowfin held in 20°C water do not become sexually active. The Pacific bonito inhabits waters off the west coast of the Americas in the northern hemisphere from south eastern Alaska (60° N) to the Baja peninsula (20° N) and in the southern hemisphere from the equator to central Chile (35° S). Sarda chiliensis is replaced by another species (Sarda orientalis) in the tropical waters west of Central America. S. chiliensis is further classified into northern and southern subspecies (S. chiliensis lineolatus and S. chiliensis chiliensis respectively). Bonito attain sexual maturity after two years of age and spawning season coincides with summer months as the water temperatures rise (Collette and Nauen 1983). Goldberg and Mussiett (1984) and Barrett (1971) found the smallest mature females of S. c. chiliensis ranged between 48.5 cm and 51.0 cm standard fork length (SFL). Bonito are known to spawn in captivity (Magnuson and Prescott 1966) and were examined in this study because they exhibited spawning behavior and eggs have been found in filters of the Monterey Bay Aquarium (MBA) and the Tuna Research and Conservation Center (TRCC). The major objective of this study was to perform a comparative examination of scombrid reproductive biology using species with distinct behavioral characteristics both in the wild and in captivity. The goal was to use the records of captive fish as a tool to learn the stages of sexual maturity in all scombrids. The assumption was that characteristics of reproductive histology are conserved among the different members of this family. I coordinated both gross anatomical and histological observations, length/age/weight regressions, and gonosomatic indices, in order to give a preliminary assessment of the best technique to determine a mature fish. Using this information, tagging projects will know more conclusively what size class to target and how to interpret data received back from different sizes of fish. Materials and Methods Tissue Sampling Tissue samples from yellowfin were taken from fish held in captivity at the Monterey Bay Aquarium (MBA) and the Tuna Research and Conservation Center (TRCC) where detailed records of their diet and water temperature had been kept as well as complete necropsy reports for all mortalities since 1995. Passive Induced Transponder (PIT) tags were used to keep track of fish. These fish had been held in 20°C water (42°0). Gonads were blotted and weighed and a small cross section was preserved in 10% formalin buffered with O.1 M cacodylate. From these tanks, 54 males and 72 females had complete information on sex, length, body mass, and gonad mass. A chart of the relationship between length and gonad mass for each gender was created from necropsy information and a regression line was fitted to the curve (Fig. 1). Samples representing this curve were selected along with fish that stood out as either being big or having large gonads. The total sample size for each sex was 15. In addition, 3 female yellowfin samples were analyzed from Kewalo Basin, Hawaii, where similar size fish (ranging from 86 cm to 92.3 cm) had also been held in captivity at warmer water temperatures. Bonito tissue samples also came from the MBA and the TRCC. Necropsy reports for the past bonito mortalities were not as detailed as the reports for the tuna, therefore this collection was limited to mortalities occurring during the course of this study. Bonito that were killed in a separate physiological experiment had gonads sampled for this study. Five females and 7 males were examined (for sizes see Table 1). Upon death, gonads were extracted and weighed. A cross-section from one gonad was preserved in 10% formalin buffered with 0.1 M cacodylate and stored at 4° C. Another, larger cross-section was frozen in liquid nitrogen and stored in a VWR brand 50 mL plastic falcon tube at-80°C. One of the two gonads was placed in a Zip-lock plastic bag and refrigerated for later examination. The samples of wild Atlantic bluefin were taken from a variety of locations. Gonad samples for histology were taken from purse seine vessels in New England (n-99 during the fall, sport fishing boats in Cape Hatteras, North Carolina (n-16) in the winter, longlines in the Gulf of Mexico (n-4) in March 1999, and by sport fishermen in the Bahamas (7-2) in April 1999. Gonad samples from the Bahamas came from an unrevealed source because they were caught illegally. They were first preserved in gin by the fishermen and then sent to Dr. Barbara Block via NMFS officials. These samples were then transferred to 10% formalin buffered with 0.1 M cacodylate. All other samples were collected by the staff of the TRCC and various sport fishermen and gonad cross-sections were preserved in 10% formalin buffered with 0.1 M cacodylate. All histological tissues were embedded in parrafin, sectioned at 10 um and stained with hemotoxylin and eosin. GSI Gonosomatic Indices (GSI's) represent the percentage of overall body weight that is gonad weight and were calculated using the following equation: [(gonad mass) (body mass))' 100. This gave a rough indication of the state of maturity. The GSI's were calculated for every sampled fish with enough data and then plotted against the length of the fish (Fig. 2, 3). Generally the wild samples did not have gonad mass data so GST's could not be calculated. Histological Classification Sexual maturity and spawning status was determined histologically according to the methods outlined by Schaefer (1998). This system of classification was derived from earlier methods developed on related fish such as northern anchovy Engraulis mordax, and skipjack tuna, Katsuwonus pelamis and have been used on southern bluefin (Farley 1998) and a variety of species within the scombrid family. All tissues were examined under an Olympus BH-2 light microscope with an Olympus DPIO digital camera attachment. Females Ovaries are considered asynchronous, meaning multiple stages of oocytes can be present in the ovary at once and that fish are capable of multiple spawnings per season. (Schaefer 1998). Females were staged based on the most advanced oocytes present in the ovary. Oocytes were classified into five stages: (1) unyolked, (2) early yolked, (3) advanced yolk, (4) migratory nucleus, (5) hydrated (Fig. 4A,B,C; 6A). Postovulatory follicles were classified into three stages depending on time since ovulation: O h, 512 h old, and 212 h old. Postovulatory follicles are only identifiable up to 24 h after ovulation before they are completely resorbed. Yolked oocytes that had not been ovulated and were being resorbed were classified as atretic. Two stages of atresia, a and B were distinguished (Fig. 4E, F) to describe early (G) atresia where the yolk is being resorbed, and late (B) atresia when the thecal and granulosa cell layers are rearranged and resorbed (Farley 1997, Schaefer 1998). The relative frequency of these stages in the ovary provided the criteria to estimate sexual status. Females were put into one of the following categories 1) Immature- Only unyolked oocytes present. No signs of atresia. 2) Mature, nonspawning- Advanced yolked or atretic oocytes are present but no sign of imminent or past spawning (migratory nucleus, hydrated, or postovulatory follicle). 3) Mature, spawning- Advanced yolked oocytes are present as well as signs of spawning (migratory nucleus, hydrated, or postovulatory follicle). Less than 100% of advanced yolked oocytes are in a atresia. 4) Postspawning- Either: 1) 2 50% of both early and advanced yolked oocytes are in o atresia; 2) 100% of advanced yolked oocytes in the a stage of atresia; 3) no yolked oocytes are present but oocytes in the ß stage of atresia are, and residual hydrated oocytes may or may not be present. Males Testes are considered an unrestricted spermatogonial type, meaning the distribution of spermatogonia and stages of spermiogenesis occur along the entire lobule Testes were classified depending on stages of spermiogenesis in the lobule walls and presence or absence of sperm in the lobular lumen and sperm duct. Three groups were adapted from Schaefer (1998) and Grier (1981): 1) Immature- Lobules relatively thin with a homogeneous stage of cysts in early development (spermatogonium). A minimal amount of variety in the stages of spermiogenesis may be present. Little or no spermatozoa in lobular lumen and central longitudinal sperm duct (vas deferens). 2) Mature, Inactive- No prior description existed because only seen in bluefin. Very similar to immature. Spermatozoa present in the lobular lumen. Lobules small and narrow surrounded by 2 30% somatic tissue. 3) Active-Lobules large with thick, darkly staining walls representing all stages of spermiogenesis. Lumen and vas deferens packed with spermatozoa. 4) Spawning- Same as mature classification but vas deferens empty. Many of the histological samples were not complete cross-sections and therefore the vas deferens was not available to determine between non-spawning and spawning mature fish. In this case the lobules were used to determine if the fish was active or not and the sample was noted as incomplete. Results GSI Female yellowfin held in captivity at the TRCC seemed to show a great deal of variation in the GST's as well as a general trend for higher GSI's with greater lengths (Fig. 2B). When compared to data from wild fish however, all values were significantly low. Schaefer (1998) reports that sexually active fish have GSI's of one or higher. None of the female yellowfin in captivity had GSI's over one. Male yellowfin GSl's were also low (Sl) and did not show any obvious trend (Fig. 2A). Bonito GST's on the other hand, were much higher and showed a great deal of variation. Females all had GSI's greater than one and a single individual was as high as 4.85 (Fig. 3B). Comparing these numbers to the histology, I found that a GSI between 1 and 2 indicates a mature but sexually inactive fish, a GSI of about 2 or above indicates sexually active but not spawning, and a GSI of 4 or higher indicates a spawning bonito. Male bonito had GST's ranging from 1.85 to 5.85 (Fig. 3A). All were sexually active but did not appear to be spawning. Histology Yellowfin Examination of histology slides from the captive female yellowfin show that none of the fish were sexually active. Curved lengths ranged from 54.3 cm to 121 cm with 9 out of 15 over 92 cm (Schaefer's length at which 50% are mature). All oocytes were in the unyolked stage and ranged in diameter from 20-120 um (Fig. 5A,B). There was a slight correlation between gonad size and average oocyte diameter even though the oocytes were all the same stage. The 3 samples from Kewalo Basin were similar except for the biggest fish (92.3 cm). It contained a low frequency of slightly larger oocytes (140 um) with grainy cytoplasm, and light colored nuclei surrounded by lipid droplets (Fig. 4B). They appeared to be between unyolked and early yolked stage. Some possibly atretic oocytes were also present in low frequency (Fig. 4E). Captive male yellowfin ranged from 53.7 cm/1 g to 113 cm/44 g and 105 cm/103.8 (curved length/ gonad mass). From the sub-sample (7-15) 12 fish were longer than the length at which 50% are mature (69.0 cm, Schaefer, 1998) and 6 were longer than the length at which 90% are mature (89.8 cm). At least 9 fish were mature and active (Fig. 7C), the rest were classified as immature with various stages of teste development (Fig. 7A). No male fish were examined from warmer waters. Bonito Curved lengths for female samples ranged from 57 cm to 65 cm, all larger than the smallest mature fish found by Barrett (1971) and Goldberg and Mussiett (1984), 51 cm and 48.5 cm respectively. One of the female bonito was classified as sexually inactive and 3 as sexually active (7—4). Only one of the active fish was classified additionally as spawning. This ovary included advanced yolked and atretic oocytes (Fig. 4D,E,F) as well as some hydrated oocytes (Fig. 6A) indicating imminent spawning, however, the low frequency of hydrated stage oocytes and the absence of the migratory nucleus stage suggest either past spawning (more than 24 h ago, allowing the follicles to be completely resorbed) or a narrow window of time prior to spawning in which hydrated but not migratory nucleus stage are apparent. Male bonito were all mature (7-5) but not spawning (Fig. 7C). Curved lengths ranged from 61 cm to 67 cm which were greater than the smallest reproductively active male (39 cm) found by Goldberg and Mussiett (1984). Past spawning in males can only be determined 12 h after the event because the vas deferens refills with spermatozoa almost immediately (Schaefer 1998). Bluefin Female bluefin from New England (7-38) ranged from 193 cm to 264 cm (curved fork length). None of the fish sampled were sexually active and the ovaries contained oocytes averaging 65 um in diameter (ranging from 20-120 um) and all in the unyolked stage of development. A few fish contained a low frequency of atresia suggesting that these fish were mature, however, this resorption was considered incidental and not indicative of sexual activity. Males ranging from 200 cm to 274 cm (n-62) also showed no signs of sexual activity. No signs of active spermiogenesis were apparent and the lobular lumen were completely devoid of sperm (Fig. 7B). Female bluefin from North Carolina (7-5) ranged from 163 cm to 240 cm. Some fish did not have either curved length or mass recorded. Missing data was filled in using a regression from NMFS (unpublished data). None were classified as sexually active. Due to poor histology, differences between immature and mature (inactive) could not be distinguished. One ovary contained a very low level of a atresia but no sign of yolk development. This atresia was considered incidental and not indicative of sexual activity. Males (7=5) were all sexually inactive as well. Female bluefin from the Gulf of Mexico (n-4) ranged from 217 cm to 239 cm. Each fish represented a successive stage of ovary development. One fish (217 cm) had an ovary containing large unyolked oocytes including the intermediate stage prior to vitellogenesis (Fig. 4B) and a low frequency of incidental atresia but did not appear to be sexually active. The other three all were sexually active but with subtle differences of ovarian development. One fish (220 cm) had ovaries containing early yolked oocytes (Fig. 40) and a low level of atresia. Another fish (239 cm) had ovaries containing more advanced yolked oocytes (Fig. 4D) but still very low levels of atresia. The last fish (218 cm) had all stages of vitellogenesis present in abundance as well as a high degree of o and ß atretic oocytes (Fig. 4). This was considered the most developed ovary. It is interesting to note that the fat pad associated with the gonad was reported resorbed and the gonadal tissue was highly vascularized. No males were sampled from the Gulf of Mexico. In the Florida Straits, one female bluefin was caught and measured to be 673 lbs (est. = 263 cm). Its ovaries contained an abundance of advanced yolked oocytes, & level atresia and very few unyolked oocytes. One 916 lb male, approximately 290 cm, was also sampled and the testicular lobules packed with spermatozoa and the lobular walls were darkly stained and thick with cysts in different stages of spermiogenesis. Both male and female were sexually active. Behavior Behavioral data collected by a PSAT tag allowed further analysis of what I saw in the histology from the Gulf of Mexico. On the same longline set that caught three of the four fish sampled histologically, a live fish was tagged and released. Approximately two months later it popped up and started transmitting data back to the TRCC. The tag came up about 90 miles from where it was released. Depth and temperature recordings showed normal tuna behavior (moving up and down through the water column and consequently encountering a variety of water temperatures) from March 21 to May 9 (Fig. 10). From May 9 until the pop up date (May 20), depth records show that all of the fish's time was spent within the top 10 m below sea level and within a narrow temperature range between 25°C and 27.5°C (Fig. 11). Discussion Although the basic reproductive biology of tuna remains relatively consistent among the different species, important differences in behavior and life history lead to unique characteristics of each. While the similarities allow us to learn about the unknown stages of less known species, the differences keep the investigation interesting and crucial for properly managing respective fisheries. Female tuna, even if mature, will not become sexually active unless the environment meets specific requirements. These have been speculated to include food availability, salinity and most importantly temperature. Females of most species of tuna will only become sexually active (start developing the ovaries) if the SST is consistently 25°C or warmer (Sund, 1981; Schaefer 1998). Bonito, an ectothermic sister taxa to tunas, become sexually active and spawn in temperatures as cold as 20°C. While warmer water temperatures do correlate to an increase in percentage of sexually active fish for S. chiliensis (Goldberg and Mussiett), the lower range of temperature at which bonito can become sexually active makes sense considering their geographic distribution in the higher latitudes. It is unknown if an upper temperature limit exists for spawning in S. chiliensis, but if so, it might explain the gap in their distribution across the tropics. The natural range of the yellowfin tuna includes the tropical Pacific where SST is consistently high. In this region, some spawning occurs year-round with peak activity in direct correlation with SST (Schaefer, 1998). As shown in this study, female yellowfin do not spawn, nor do they become sexually active when held in 20°C water. Other factors hypothesized to have an effect on sexual activity (such as food availablility and salinity) are apparently less important factors. I found that even though diet has been variable over the four years of keeping female yellowfin in tanks, even the samples that had been fed a high fat, high calorie diet do not become sexually active. Males on the other hand seemed to show no response to the environmental conditions around them. Even when fed the current low calorie, low fat diet at the TRCC, sexual activity coordinated with the expected sizes at maturity. Wild bluefin tuna appear to show similar responses to SST. Ovaries from the northern fish in colder waters regress to an inactive state and are practically indistinguishable from immature fish. Occasionally, a low frequency of incidental atretic oocytes are apparent which may be the only way to determine the difference between an immature and a mature but inactive bluefin in the temperate part of their range. This technique is not dependable though, since such a low frequency of resorbing oocytes means that a 10 um cross-section may not include this stage. Consequently female bluefin from New England and North Carolina appeared similar to the tuna kept in cold water at the TRCC. Male bluefin from these regions looked like the immature yellowfin at the TRCC. The difference between male and female yellowfin is perhaps due to the relative investment of energy of each sex. Schaefer (1998) estimated that a female yellowfin in the wild, releases approximately twice her body weight in spawn each year. It would be detrimental to release such a large investment into conditions that are not conducive to egg and larval growth. Males however, don’t have as much invested in their gonads and can afford to have them ready to spawn at all times. The surprising result of this study is the difference between yellowfin males and bluefin males. In contrast to yellowfin, bluefin males are not sexually active when in the colder part of their range. This suggests that neither temperature nor food availability play a role in male sexual development. Rather, yearly cycles based around customary migration patterns seem to be more reasonable as determining factors for sexual state. If the hypothesis that only mature bluefin come to the Gulf of Mexico and they are there to spawn is correct (Nemerson 1998), the most intuitive way to determine the size of sexual maturity is to look at the size of the fish arriving in the Gulf. Size-catch statistics have been collected on the Japanese longline industry and by NMFS landings for more than 30 years. According to this data, few fish (0.4 percent, Nemerson et al. 1998) are caught during the breeding season under approximately 135 kg and 185-200 cm (Fig. 9) This size frequency is consistent even before 1982, when a minimum size limit for the retention and sale of small bluefin was established which may have led to an under- reporting of smaller fish. It is possible that due to the limited extent of bluefin targeted fishing, smaller fish were missed. Considering the more than thirty year history of the fishery and the parallel fishery for yellowfin which targets smaller fish, the idea of a large representation of small bluefin going completely unnoticed is unlikely. Therefore, based on spawning ground catch statistics alone, the first length at which a majority of western Atlantic bluefin tuna are mature and able to spawn is 200 cm, correlating to an age of approximately 9-10 years. However, if the assumptions prove to be false, if immature bluefin can show up in the Gulf of Mexico or if another spawning ground exists, the length at maturity could increase or decrease respectively. The four samples from the Gulf of Mexico and one from the Bahamas that I examined all represented different stages of development towards sexual activity. Even on the spawning grounds, in large fish (215 cm), not all were sexually active. A female bluefin (218 cm) from the Gulf of Mexico appeared sexually inactive and therefore I could not confidently determine if it was mature given only the histology. The following information about this fish suggests it was mature and becoming sexually active. First, its size correlated with the other active fish which represented the same size/age class and were caught on the same longline. This could mean a great deal of variation exists in length at maturity, or in the time it takes for a fish to switch from an inactive to an active state. Related to the latter possibility is the time of year at which these fish were captured. All bluefin from the Gulf of Mexico were taken in March, the supposed beginning of their spawning season (Baglin, 1982; Rivas 1964). Most likely, they had just arrived from the colder feeding grounds and were beginning to respond to the new environment. Studies on related fish have shown the tendency of older and bigger fish to start spawning earlier in the season and be more fecund. This theory is questionable due to the vast differences of sexual status of three fish from the same size/age class. Secondly, the tag data recovered from a cohort of the fish I examined histologically, suggests that it waited approximately 2 months before starting to spawn. I hypothesize that the drastic change in behavior represents spawning behavior because 1) most of its cohorts were sexually active and almost ready to spawn in the beginning of the spawning season, and 2) spawning would most likely occur in the warmest waters which is where this fish started spending 100% of its time. The data transmission was incomplete and many days were either incomplete or missing altogether. Nevertheless, the surface pattern was distinct and long enough that it stood out clearly. Taking into account the new data presented in this study, as well as the results from previous studies, the Gulf of Mexico and the Florida Straits still appear to be the most significant spawning ground for the western Atlantic bluefin tuna. While this has historically been considered the primary site, recent controversy has risen over the speculation of alternative spawning sites (Lutcavage, 1999). My analysis shows alternative sites are very unlikely. Although sample sizes were low in some places, the lack of any evidence of sexual activity, let alone spawning, suggests that western bluefin spawning grounds are confined to the Gulf of Mexico and the Florida Straits. More sampling is needed from different seasons and if possible, from the new sites proposed as alternatives. My comparative analysis of the males provides some additional evidence regarding the factors determining sexual activity. In the 20°C tank water, male yellowfin matured and became sexually active at sizes equivalent to those found in the wild according to the study done on wild yellowfin (Schaefer, 1998). Bluefin provide a contrast to this by being sexually inactive in the northern part of its range where water temperatures are colder. Testes can regress quickly, making the identification of mature males difficult. Male yellowfin seem to have a different sexual response to colder water temperatures than the females. Male responses to water temperatures are variable between species. Unlike females, males probably have other stimuli to which they respond. Perhaps these stimuli could come from the presence of active females, food availability, yearly sexual cycles or a genetically inherited trait that has been developed along with behavioral characteristics. The mechanism is purely speculative at this point. Female bluefin must either have consistently warm water to begin development, or the time of vitellogenesis is pre- programmed. Males require either similar conditions or the presence of active female. Any of these hypotheses make it unlikely for bluefin to spontaneously spawn wherever or whenever the appropriate conditions are met. Just because the environment is temporarily correct for spawning in a gulf stream eddy, it does not mean bluefin will spawn there. Conducive spawning conditions must either be consistent, familiar, or both. Knowing the amount of time it takes for each sex to go from the mature, inactive state to an active one would be a great help to determine where spawning is possible. Assuming no heritable influence in yearly cycles, how long after arriving in the Gulf, or a region conducive to spawning, will fish start to become sexually active and eventually spawn? If the behavior recorded by the PSAT in the Gulf of Mexico does represent spawning behavior, then this bluefin took at least two months to start spawning. This data also suggests that, as previously hypothesized, the peak spawning season is in May when SST’s are most conducive for larval devlopment. More tags back from this region may reveal the degree of variability within the species. In conclusion, the gonad histology of scombrids is very similar and the stages of sexual maturity of one species can be used to identify the stages of another. Yellowfin held in captivity, in water temperatures below what they usually encounter in the wild, do not spawn. Female yellowfin do not become sexually active in water temperatures of 20°C but males do, with no apparent deviation from the size ranges found in the wild. Bonito become sexually active and spawn continuously when held in captivity in the same water temperatures (20°C). Atlantic bluefin (both male and female) found in the northern part of their range are in a sexually inactive state and only become active when they migrate to the spawning grounds. Only fish 200 cm and over make this migration. Upon arrival, a certain amount of time is required to develop into a spawning state. This time period is unknown but recent tag data shows it may be as long as two months. In the future, the captive fish can be used in further sexual development experiments by altering the water temperatures and diets in the tanks. Histologically, wild sample sizes can be increased and taken from different regions at different seasons. Finally, more tagging data is still being recovered and these results should elucidate the questions of spawning site fidelity, mixing between stocks and migratory habits of the giant bluefin tuna. Acknowledgements thank my advisor Dr. Barbara Block, as well as the great team of people she has brought together, for their enthusiasm and advice. I thanks NMFS for providing data on Gulf of Mexico landings. I also thank Chuck Farwell of the Monterey Bay Aquarium and Dr. Tom Williams for their encouragement, and finally, Andy Seitz for assistance in day to day tasks as well as an insight into the world of tuna. References Baglin, R.E. 1982. Reproductive biology of western Atlantic bluefin tuna. Fish. Bull. 80:121-134. Block, B.A., H. Dewar, C. Farwell, and E.D. Prince. 1998. A new satellite technology for tracking the movements of Atlantic bluefin tuna. Proc. Natl. Acad. Sci. U.S.A. 95:9384-9389. Carey, F.G. and J.M. Teal. 1969. Regulation of body temperature by the bluefin tuna. Comp. Biochem. Physiol. 28:205-213. Collette, B.B. and C.E. Nauen. 1983. FAO species catalogue. Vol. 2. Scombrids of the world. An annotated and illustrated catalogue of tunas, mackerels, bonito and related species known to date. FAO Fish. Synop. 125, Vol. 2:51 Farley, J. H. and T.L.O. Davis, 1998. Reproductive dynamics of southern bluefin tuna, Thunnus maccoyii. Fish. Bull. 96:223-236. Golderg, S.R. and D. Mussiett. 1984. Reproductive cycle of the Pacific bonito, Sarda chiliensis (Scombridae), from northern Chile. Pacific Science. Vol. 38, no 3:228-230. Grier, H.J. 1981. Cellular organization of the testis and spermatogenesis in fishes. Am. Zool. 21(2):345-357. Lutcavage, M.E., R.W. Brill, G.B. Skomal, B.C. Chase, and P.W. Howey. 1999. Results of pop-up satellite tagging of spawning size class fish in the Gulf of Maine: do North Atlantic bluefin tuna spawn in the mid-Atlantic? Can. J. Fish. Aquat. Sci. 56:173-177. Magnuson, J.J. and J.H. Prescott. 1966. Courtship, locomotion, feeding, and miscellaneous behaviour of Pacific bonito (Sarda chiliensis). Anim. Behav. 14:54-67. McGowan M.F. and W.J. Richards 1989. Bluefin tuna, Thunnus thynnus, larvae in the Gulf Stream off the Southeastern United States: Satellite and shipboard observations of their environment. Fish. Bull. 87:615-631. Nemerson, D., S. Berkeley, and C. Safina. In press at Fish. Bull. Patinno, Reynaldo. 1995. Gonads. pp. 128-153 in Takashima, F. and T. Hibiya, eds. An Atlas of Fish Histology, Second Edition, Gustav Fischer Verlag, Germany Rivas, L.R. 1954. A preliminary report on the spawning of the western north Atlantic bluefin tuna (Thunnus thynnus) in the Straits of Florida. Bull. Mar. Sci. Gulf Carib. 4:302-322. Schaefer, K.M. 1996. Spawning time, frequency, and batch fecundity of yellowfin tuna, Thunnus albacares, near Clipperton Atoll in the eastern Pacific Ocean. Fish. Bull. 94:98- 112. Schaefer, KM. 1998. Reproductive biology of yellowfin tuna (Thunnus albacares) in the easter Pacific Ocean. Inter-American Tropical Tuna Commission Bulletin. Vol. 21, No. 5:205-272. Sund, P.N., M. Blackburn, and F. Williams. 1981. Tunas and their environment in the Pacific Ocean: a review. Oceanogr. Mar. Biol. Ann. Rev. 19:443-512. Figures Fig. 1) A. Gonad mass compared to curved length of captive female yellowfin tuna held at 20°C. B. Gonad mass compared to curved length of captive male yellowfin tuna held at 20°C. Fig. 2) Gonosomatic indices for captive male yellowfin (A) and captive female yellowfin (B) plotted against curved length with linear trend line added. Fig. 3) Gonosomatic indices for captive male bonito (A) and captive female bonito (B) plotted against curved length with linear trend line added Fig. 4) Various stages of oocyte development found simultaneously in a bluefin tuna ovary. A. Unyolked (a) stage of oocyte development. Bar - 0.1 mm. B. Slightly more advanced stage (a) of unyolked oocyte prior to yolk development. Bar - 0.1 mm. C. Early yolked (a) stage of oocyte development. Red granules represent yolk, white represents lipid globules. Development of the vitelline envelope is also apparent. Bar- 0.1 mm. D. Advanced yolk (a) stage. Bar - 0.1 mm. E. Unovulated yolked oocyte in the alpha stage of atresia (a). Bar - 0.1 mm. F. Vitellogenic oocyte (a) occurring with oocyte in the beta stage of atresia (b). Bar - 0.1 mm. Fig. 5) A. Immature ovary from a small yellowfin tuna (64.5 cm) at the TRCC, with a high percentage of somatic tissue (a), and all small unyolked oocytes (b). Bar - 0.1 mm. B. Mature, inactive ovary from a large captive yellowfin (105.5 cm), with larger unyolked oocytes (a), and a small degree of incidental atresia (b). Bar - 0.1 mm. C. Sexually active ovary from a giant bluefin (217 cm)with multiple stages of oocyte development present simultaneously. Unyolked (a), Early yolked (b), Advanced yolked (c), Alpha atresia (d), and Beta atresia (e). Bar - 0.1 mm. Fig. 6) A. Hydrated oocyte found in female bonito with the highest GSI. Indicates imminent spawning. Bar -O.1 mm. B. Female bluefin from the Florida Straits with advanced yolked oocytes (a). Represents first histological evidence of bluefin sexual activity outside of the Gulf of Mexico. Photograph taken at same power magnification as above (A). Fig. 7) A. Cross section of lobules in an immature testis with thin, lightly stained walls (a), lumen devoid of spermatozoa (b), and high percentage of somatic tissue. Found in a small yellowfin (61 cm). Bar - 0.1 mm. B. Cross section of lobules in a mature, sexually inactive testis, from a giant bluefin (220 cm) in New England, also with lightly staining walls (a) and empty lumen (b). Bar - 0.1 mm. C. Cross section of lobules in a mature, sexually active testis taken from a bonito with high GSI. Dark staining walls with various stages of spermiogenesis (a) and lumen packed with spermatozoa (b). Fig. 8) Size and sex distributions of wild bluefin sampled for this study according to region. New England (A), North Carolina or Mid Atlantic Bight (B), and Gulf of Mexico and Florida Straits (C). Fig. 9) National Marine Fisheries Service observations of the Japanese longline fishery in the Gulf of Mexico over the last five years (1995 to 1999). Fig. 10) Depth and temperature data from a pop up satellite archival tag (PSAT) for May 5, 1999. Readings were taken every 60 seconds and grouped into four hour blocks (6 per day). A. Depth categories go down the y-axis of with the surface on top. Percentage of time during a 4 h block is indicated along the x-axis. Different series represent the different 4 h blocks during the course of the day. The gradient color scheme goes from lightest (earliest) to darkest (latest). B. Temperature categories are scaled down the y- axis with the hottest at the top. Similar color gradients also represent time of day. Fig. 11) Depth and temperature data for May 17, 1999 using the same format as fig. 10. Table 1 A summary of the histological results. Region Sex Species Bonito TRCC TRCC TRCC Yellowfin TRCC Bluefin N.E. N.E. N.C. N.C. G.Of M. G.Of M. F. Straits F. Straits M Size Range (cm 57-65 61-67 54-1 53-113 193-264 201-274 158-215 115-240 217-239 217 239 Immature Mature 0 Active 0 0 0 . 28 Spavon A. B. 250 200 150 100 50 40.0 120 100 80 60 40 20 40.0 Captive Female Yellowfin Tuna y=0.05162-6.0959x + 184.3 R2=0.9005 8 60.0 80.0 100.0 Curved Length (cm) Captive Male Yellowfin Tuna y=0.014x - 1.5596x + 45.151 R?=04257 80.0 60.0 100.0 Curved Length (cm) 120.0 * 140.0 120.0 Figure 1. Figure 2 30 Captive Male Yellowfin Tuna ------------ ---------------------------------------------- 0.6 --------------------------------------- 0.5 —-------+------------- 04 -- 03 y=0.0019x-0.045 R?- 00542 --------- --------------- -------- ------------- 0.2 28 0.1 +————— 85 40 80.0 90.0 100.0 110.0 120.0 40.0 50.0 60.0 70.0 Curved Length (cm) Captive Female Yellowfin Tuna 0.9 7 9 0.7 +- ------- 0.6 + 0.5 +0— - 04 -------------- --------------- 0.3 +- 2 9 -------------- — 0.2 + y=0.0096x-0.4675 R2-0.5904 ————— 0.1+- 0 + 400 50,0 600 70.0 80.0 90.0 100.0 110.0 120.0 130.0 Curved Length (cm) Figure 3 31 Male Bonito 71 6-- 5 4 5 2 -------------------------- ----- -------- 56 68 70 Curved Fork Length (cm) Female Bonito --------------------- 4 +----------------------------------------------------------------------- 3 +------------------------------------------------------------------------------------ 2Q ----------------- -------- ------ Curved Fork Length (cm) Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 25 20 15 New England I EE 240 250 260 270 280 290 190 200 210 220 230 Curved Length (cm) North Carolina Hn HE II I 5150 160 170 180 190 200 210 220 230 240 Curved Length (cm) Gulf of Mexico I m m 150 160 170 180 190 200 210 220 230 240 250 260 270 280 290 Curved Length (cm) HFemale EMale Total 88 8 8 38 S8 3 o 1 8 8 ------------------------- a 3 8 8 Frequency 8 Figure 10 Percentage of Time Spent at Depth: May 5, 1999 -im -5 to 10m 10.5 to 50m 50.5 to 100m 100.5 to 150m 150.5 to 250m 250.5 to 350m 80 20 40 60 100 Percentage of Time Spent at Temperature: May 5, 1999 30.05 to 32.50 25.05 to 27.500 20.05 to 22.500 15.05 to 17.500 10.05 to 12.500 40 80 100 20 Percent Time m 20.00h E16:00h E12:00h H08:00h H04:00h Hoo:o0 h E 20.00h E 16:00h E12:00h 08:00h D04:00h Hoo:o0h Figure 11 Percentage of Time Spent at Depth: May 17, 1999 -im -5 to 10m 10.5 to 50m p; 50.5 to 100m § 100.5 to 150m 150.5 to 250m 250.5 to 350m 40 80 100 20 Percent Time Percentage of Time Spent at Temperature: May 17, 1999 30.05 to 32.50 27.55 to 30.000 25.05 to 27.500 22.55 to 25.000 20.05 to 22.500 17.55 to 20.000 15.05 to 17.500 12.55 to 15.000 10.05 to 12.500 40 80 100 20 Percent Time E 20.00h E16:00 E 12:00 08:00 E04:00h oo.o0h E 20.00h E 16:00h 12:00 Ho8:ooh 04:00 oo:ooh