2
Historical Evidence for Stock Structure

INTRODUCTION

This chapter summarizes available biological evidence on the stock structure of Atlantic bluefin tuna. The chapter reviews published and unpublished information on the issue of whether this information supports the existence of separate eastern and western management units of bluefin tuna in the North Atlantic Ocean. Included are sections on stock concepts, history of stock and population designations in Atlantic bluefin tuna, genetics, life history, climate, and movement. Except for the first two sections (on stock concepts and historical stock designations in Atlantic bluefin tuna), the term ''management unit(s)" is used interchangeably with eastern and western stocks of Atlantic bluefin tuna and the term population is used to indicate a genetically-distinct group of fish. In addition, the term "movement" is used to indicate mixing, migration, or both, as they refer to individuals.

CONCEPT OF STOCKS

Geographic boundaries of species are influenced by environment, suitable habitats, and historical events. A fish stock can be defined as all fish belonging to a given species that live in a particular geographic area at a particular time. The area can be constrained by geographic or oceanographic features (e.g., bays, temperature discontinuities), but also may be defined by political boundaries. Political boundaries are commonly used in fisheries management, but a stock defined in this way generally will not reflect biologically meaningful management units.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 9
An Assessment of Atlantic Bluefin Tuna 2 Historical Evidence for Stock Structure INTRODUCTION This chapter summarizes available biological evidence on the stock structure of Atlantic bluefin tuna. The chapter reviews published and unpublished information on the issue of whether this information supports the existence of separate eastern and western management units of bluefin tuna in the North Atlantic Ocean. Included are sections on stock concepts, history of stock and population designations in Atlantic bluefin tuna, genetics, life history, climate, and movement. Except for the first two sections (on stock concepts and historical stock designations in Atlantic bluefin tuna), the term ''management unit(s)" is used interchangeably with eastern and western stocks of Atlantic bluefin tuna and the term population is used to indicate a genetically-distinct group of fish. In addition, the term "movement" is used to indicate mixing, migration, or both, as they refer to individuals. CONCEPT OF STOCKS Geographic boundaries of species are influenced by environment, suitable habitats, and historical events. A fish stock can be defined as all fish belonging to a given species that live in a particular geographic area at a particular time. The area can be constrained by geographic or oceanographic features (e.g., bays, temperature discontinuities), but also may be defined by political boundaries. Political boundaries are commonly used in fisheries management, but a stock defined in this way generally will not reflect biologically meaningful management units.

OCR for page 9
An Assessment of Atlantic Bluefin Tuna Criteria for managing fishery stocks under the Endangered Species Act of 1973 (ESA) have been discussed by Waples (1991) and under the Magnuson Fishery Conservation and Management Act of 1976 (MFCMA) in a recent National Research Council report (NRC, 1994). The MFCMA (Public Law 94-265, 16 U.S.C. 1801 et seq.) specifies that "an individual stock of fish shall be managed as a unit throughout its range," but does not provide criteria for defining a stock. The original ESA also did not provide criteria for defining a stock or management unit, but the ESA amendments of 1978 (Public Law 95-632 [1978], 92 Stat. 3751) defined a "species" as "any subspecies of fish or wildlife or plants, and any distinct population segment of any species of vertebrate fish or wildlife which interbreeds when mature." In the past few decades, biochemical and molecular genetic methods have been applied to fishery management issues, leading to an expansion of the stock concept to include interpopulation genetic variability. This expanded stock concept is intended to facilitate the conservation of biologically meaningful management units (Utter, 1981; Waples, 1991) that may be uniquely adapted to a particular area. Waples (1991) proposed that a distinct population segment should be defined as an "evolutionarily significant unit" (ESU) that is "substantially reproductively isolated from other conspecific populations units'' and that "represents an important component in the evolutionary legacy of the species." Under this concept, populations are defined as groups of individuals that share a common space, interbreed, and are totally or partially isolated from other such groups. The degree of isolation, brought about by reduced gene flow among breeding areas and the amount of time the populations have been isolated from one another, influences the degree of genetic differentiation among groups. Dizon et al. (1992) proposed a scheme of population classification based on genetic and geographic criteria. Category I populations are geographically separated groups of individuals that are more closely related genetically to each other than they are to individuals in other groups. Such populations have the highest probability of being evolutionarily significant units. Category II populations are also differentiated genetically but are only marginally separated geographically. Category III populations show little genetic differentiation from one another but are geographically separated and therefore likely to be isolated reproductively. Genetic differences among geographically isolated populations are expected eventually to increase. Category IV populations show little genetic differentiation because of extensive gene flow and are unlikely to be evolutionarily significant units. Geographic distribution, parasite markers, microconstituent analysis, tag-recapture data, population parameters, morphological variability, and genetic information can be used to assign populations to these categories. A variety of evidence is available for assessing the population structure of Atlantic bluefin tuna, but much of the data are equivocal. The sections below present these data and their interpretations.

OCR for page 9
An Assessment of Atlantic Bluefin Tuna HISTORY OF ATLANTIC BLUEFIN TUNA STOCK DESIGNATIONS The two-stock hypothesis currently used by the International Commission for the Conservation of Atlantic Tunas (ICCAT) is based in part on the assumption that mixing of western Atlantic bluefin tuna and eastern Atlantic/Mediterranean bluefin tuna is limited (ICCAT, 1992, 1993). The two management units of Atlantic bluefin tuna include a western management unit (west of 45°W) and an eastern management unit (east of 45°W and in the Mediterranean Sea; see Figure 2-1). Data indicate that while spawning is limited to two discrete areas, the Gulf of Mexico and the Mediterranean Sea, there is movement of individuals between western and eastern management units. A key issue is the extent of movement. The first studies of population structure in Atlantic bluefin tuna date to the early part of this century. The earliest written work is attributed to several reports by M. Sella (1926, 1927, 1929; cited in Brunenmeister, 1980), who, in the mid-to late 1920s, inferred origins and movement patterns from fishing tackle characteristics of different eastern Atlantic Ocean and Mediterranean Sea fisheries. Sella hypothesized that bluefin tuna moved from the eastern Atlantic Ocean into the Mediterranean Sea, that tuna moved from the south of Spain to Norway after spawning, and that small and medium-sized tuna could swim long distances. Although criticized because of concerns that tackle types were not reliable indicators of hooking localities, Sella's hypotheses agree with movement patterns inferred from tagging experiments carried out since the early 1900s. An early review of research on population structure for Atlantic bluefin tuna is an unpublished report in the early 1970s by F.J. Mather and A.C. Jones (cited in Murphy, 1990), which suggested that there were three populations: one in the western Atlantic Ocean, one in the eastern Atlantic Ocean, and one in the Mediterranean Sea. They also suggested that a separate population might exist in the south Atlantic Ocean. Mather et al. (1974) suggested two alternative hypotheses: a single Atlantic population and one or more Mediterranean populations; or two Atlantic populations, one spawning in the western Atlantic Ocean and the other in the eastern Atlantic Ocean or the Mediterranean Sea, or both, and one or more Mediterranean populations. They believed there was evidence of a two-stock hypothesis but noted that the "evidence is insufficient to permit clear-cut conclusions." Brunenmeister (1980) also reviewed evidence for population structure but was unable to support any hypothesis. Finally, Murphy (1990) argued that the presently accepted two-population hypothesis was not adequately flexible to fit available data. He proposed that bluefin tuna in the northern Atlantic Ocean represented one population and that the interchange between the Atlantic and Mediterranean populations is sufficiently small such that Mediterranean bluefin tuna may represent a second population. His hypothesis appears to be based on three assumptions: (1) rates of movement between western and eastern Atlantic bluefin tuna are significantly higher than those between eastern Atlantic and

OCR for page 9
An Assessment of Atlantic Bluefin Tuna FIGURE 2-1  General distribution of bluefin tuna in the Atlantic Ocean (darkened areas indicate known spawning areas [adapted from FAO, 1968]). The solid line separates the ICCAT eastern and western management units.

OCR for page 9
An Assessment of Atlantic Bluefin Tuna Mediterranean bluefin tuna; (2) possible differences in spawning productivity exist; and (3) the Gulf Stream, via its extension in the North Atlantic Current, could transport larval and postlarval bluefin tuna into the middle of the Atlantic Ocean. GENETIC STUDIES Molecular techniques that assay genetic variability have an obvious advantage over techniques that measure life history traits where the genetic component is unknown. Such techniques can be used to quantify genetic divergence and gene flow among populations and to estimate breeding structure within populations. There are four criteria for using a molecular method to find genetic markers that may be used to define fish management units: (1) expression of the genetic markers does not change during the life of an individual; (2) barring mutation, markers are inherited unchanged from one generation to the next; (3) it is possible to assay a large number of individuals from a large number of localities to adequately resolve the genetic structure of one population; and (4) there is sufficient within-population variability to make make robust statistical tests of geographic structure. Data resulting from methods satisfying these criteria can be analyzed in several ways. If genotypic data from nuclear genes (allozymes) of randomly sampled individuals surveyed with protein electrophoresis are available, a contingency table analysis of gene frequencies can be used to test for homogeneity among different sampling localities. If significant differences are found, and one can assume migration-drift equilibrium, then one can infer that the samples were drawn from genetically discrete populations. A caveat is that large samples must be used to detect small but significant gene-frequency differences between or among areas. Another test for geographic structure is to compare the observed numbers of genotypes (AA, AB, BB, etc.) in a pooled sample with the number expected from random mating (Hardy-Weinberg proportions), AA AB BB p2 2pq q2 where p is the frequency of the A allele and q is the frequency of the B allele. If there are regional genetic differences, the pooled sample will show a significant deficit of heterozygotes (Wahlund's effect) owing to mixing of individuals from discrete populations. This test, however, lacks power to detect small but significant genetic differences between populations. This significant deficit is taken as evidence for genetic differentiation between or among populations. More recently, methods for detecting nucleotide sequence variability have been used to study fish populations. Most of this effort has been directed toward the analysis of mitochondrial DNA (mtDNA), a circular piece of DNA found in

OCR for page 9
An Assessment of Atlantic Bluefin Tuna the cytoplasm of the mitochondria outside the cell nucleus. The analysis of mtDNA is based on three important differences from the analysis of nuclear DNA. First, mitochondria are inherited only from the female parent; second, there is no recombination among mtDNA molecules, greatly simplifying interpretation of phylogenetic trees based on mtDNA; and, third, there is a higher mutation rate in mtDNA than in nuclear genes typically assayed for population discrimination, thus providing the opportunity to examine more recent divergences. These characteristics provide a finer-scale resolution of genetic differences than is possible for the analysis of nuclear genes assayed with protein electrophoresis. The rates of mtDNA evolution in some large animals, however, appear to be quite slow and the analysis of mtDNA in these animals may not provide better population resolution (e.g., skipjack tuna; see Appendix E). DNA restriction (cleaving) enzymes have been used extensively to detect mtDNA restriction site differences among haplotypes,1 and, more recently, the polymerase chain reaction (PCR) has been used to amplify specific mtDNA segments for nucleotide sequencing. One application of mtDNA restriction site or sequence information (or both) is the generation of parsimony networks for visualizing minimal mutation distances among haplotypes. These networks link mtDNA haplotypes by single gains or losses of restriction sites (or by changes in homologous nucleotides) and are superimposed on geographic localities to test whether geographic cohesion of haplotypes (or haplotype lineages) exists. If haplotypes are shared among localities, frequency distributions also can be used in contingency table or other analyses to test hypotheses of genetic homogeneity among geographic localities. The level of population structure that can be detected with a particular molecular or biochemical method depends on the mutation rate associated with the assayed DNA or its product. Techniques such as protein electrophoresis, which distinguish products of genes with mutation rates of about 10-7 per generation (Nei, 1987), are capable of detecting the effects of population events over several thousands or millions of years ago. Analysis of mtDNA, which has a mutation rate of about 10-5 per generation (Wilson et al., 1985), can potentially detect the effects of more recent population events. Recently, analysis of nuclear microsatellite loci, which appear to have a mutation rate on the order of 10-3 per generation (Valdes et al., 1993), has been used to detect events occurring on the order of thousands of years ago. Interpretation of data derived from these various methods should be tempered with an understanding of the temporal scale of population events that each method is capable, in theory, of resolving. 1    The term haplotype refers to the mtDNA genes which are inherited only from the female parent, so the mtDNA carries only a single copy of each mitochondrial gene.

OCR for page 9
An Assessment of Atlantic Bluefin Tuna Genetic Variation in Tunas and Scombroid2 Fish Bluefin tuna are members of the family Scombridae, an assemblage that contains 15 genera and about 48 species of epipelagic fish. The scombrid genus Thunnus contains seven species, including the bluefin tuna. Results of molecular genetic studies of Thunnus, as well as species that are related to bluefin tunas (mackerel, bonitos, and billfishes), have provided information on the amount of genetic differentiation that might be expected among the global populations of bluefin tuna. Several recent studies have used mtDNA sequence variation to examine patterns of molecular genetic divergence and infer evolutionary relationships among tunas and other scombroid fish. Bartlett and Davidson (1991) examined a 290-base-pair (bp) sequence of the cytochrome b (cyt b) gene in four species of tuna from the northeast Atlantic Ocean: bluefin (n = 33), yellowfin (n = 33), bigeye (n = 32), and albacore (n = 12). The main result of this study was the demonstration that one could use these 290 bp to distinguish each of the four species and that only a small amount of tissue (100 mg) is required for analysis. This was significant given that carcasses (which do not have identifying morphological characters), often end up at auctions and could be sampled and used in the genetic assessment of population structure. Also, identification of larval and juvenile specimens, which are often difficult to sort to species, is now possible with genetic techniques. In broader studies to understand the relationships among tunas, mackerels, and billfishes, Block et al. (1993) and Finnerty and Block (1994) examined 600 nucleotides of the cyt b gene among 30 species. These studies included nine species of tunas and provided direct comparison of sequence variability between northern and southern bluefin tunas. The inferred phylogeny provided strong support for the monophyly of tuna genera (Thunnus, Katsuwonus, Euthynnus, and Auxis) and indicated that species of the genus Thunnus are closely related to one another. Only three nucleotide sites differed between southern (n = 2) and northern (n = 2) bluefin tuna. Sequence differences between these populations are small (0.5%) compared to the maximum intraspecific sequence difference detected among other members of the suborder (i.e., 1.8% for blue marlin, Makaira nigricans, over a similar region of cyt b [Finnerty and Block, 1992]). These data call into question the validity of separating northern (T. thynnus) and southern bluefin tuna (T. maccoyi) into separate species or indicate the presence of Northern bluefin tuna in southern oceans. Additional molecular data, particularly from nuclear genes, are needed to determine whether the inferences made in these studies can be corroborated. In a third study of mtDNA variability in tunas, Chow and Inoue (1993) 2    The suborder that includes tunas and other fish is Scombroidei, and these fish are referred to as scombroids. When referring specifically to tuna, the family name, scombridae, is used, and these fish are referred to as scombrids.

OCR for page 9
An Assessment of Atlantic Bluefin Tuna examined restriction fragment length polymorphisms (RFLPs) of three PCR-amplified mtDNA fragments from eight species/subspecies of Thunnus : (1) cyt b (355 bp, 15 enzymes, n = 132); (2) 12S ribosomal RNA (450 bp, 20 enzymes, n = 16); and (3) part of the coding regions of the ATPase and COIII genes (ATCO, 940 bp, 20 enzymes, n = 131), as well as regions adjacent to the coding regions. Cyt b RFLP fragment differences separated these tunas into four groups. The groupings identified from the cyt b data by Chow and Inoue (1993) were discordant with groupings inferred from direct sequencing of cyt b (Block et al., 1993) and from analyses of morphological variation (Collette et al., 1984). Chow and Inoue (1993) also reported that ATCO fragments indicated differences among species. One of the 18 northern Pacific bluefin tuna had cyt b and ATCO fragment patterns identical to those of northern Atlantic bluefin tuna. The authors suggested that this reflects incomplete genetic differentiation between northern Atlantic and Pacific bluefin tuna. A more reasonable hypothesis is that the exceptional mtDNA haplotype reflects movement of Atlantic bluefin tuna into the Pacific Ocean basin. The possibility of movement of bluefin tuna between Atlantic and Pacific ocean basins merits further investigation. Genetic Variation in Bluefin Tuna Information on biochemical and molecular genetics of Atlantic (Thunnus thynnus) and southern (T. maccoyi) bluefin tuna populations is limited. One early study of frequencies of alleles coding for the protein transferrin among four samples of southern bluefin tuna in Australian waters showed significant allele-frequency differences (G = 21.02, degrees of freedom = 3, P < 0.001; Fujino and Kang 1968). In Atlantic bluefin tuna, Edmunds and Sammons (1971, 1973) found allele-frequency homogeneity at the superoxide dismutase (SOD) locus among samples from the western Atlantic Ocean (New Jersey, n = 269; Rhode Island, n = 87; and Nova Scotia, n = 25) and between these samples and one from the Bay of Biscay (n = 675) in the eastern Atlantic Ocean (G-test, P > 0.05). Edmunds and Sammons (1973) pooled these samples into a single sample (n = 1,056) and tested the pooled sample for fit to Hardy-Weinberg proportions. No deviation from expected equilibrium proportions was found. This was taken as evidence for the lack of genetic differentiation between eastern and western Atlantic bluefin tuna. The oceanwide fit of SOD genotypes to proportions expected from random mating is consistent with the hypothesis that a single population of bluefin tuna occurs in the North Atlantic Ocean. Using allele frequencies at three nuclear gene loci, Phipps (1980) tested the hypothesis that early-and late-arriving bluefin tuna in St. Margaret's Bay, Canada, did not differ genetically. Most bluefin tuna enter the bay in two waves, one in July and a smaller one in mid-September/October (429 were sampled in the first wave, 16 in the second wave). No departures from expected Hardy-Weinberg proportions within or among samples were found at SOD-1, SOD-2,

OCR for page 9
An Assessment of Atlantic Bluefin Tuna and G6PD (glucose-6-phosphate dehydrogenase). Allele frequencies at SOD-1 between the two samples did not differ significantly, whereas allele frequencies at SOD-2 (G = 12.11, P < 0.01) and G6PD (G = 137.19, P < 0.001) did differ significantly. There was evidence for a null allele at SOD and for artifacts of G6PD banding, indicating that the differences between samples must be interpreted with caution. In addition, the sample of late-arriving fishes was small (n = 16). Bartlett and Davidson (1991) sequenced 290 bp of the cyt b gene and found six mtDNA haplotype among 33 individuals of Atlantic bluefin tuna; 28 of the fish shared a single haplotype. The common haplotype differed from four others by one nucleotide substitution and from one other by two substitutions. Because the portion of the cyt b gene sequenced by Bartlett and Davidson (1991) generally is conserved in other scombroids (e.g., swordfish [Finnerty and Block, 1992]), it likely will not be informative for resolving population structure in Atlantic bluefin tuna. Other regions of the mtDNA molecule (e.g., the D-loop) are more variable and thus may be more suitable for resolving population structure in bluefin tuna. To date, molecular genetic studies of bluefin tuna have not focused on the issue of genetic divergence among global samples of bluefin tuna. The genetic analysis of within-ocean basin diversity of Atlantic bluefin tuna (e.g., Atlantic Ocean, Mediterranean Sea), would benefit from a worldwide study of molecular genetic variation among bluefin tuna. Molecular genetic studies in other highly migratory, scombroid fishes (tunas, marlins, and swordfish) have demonstrated the utility of such an approach (Appendix C), and a thorough analysis of nucleotide sequence variability in both mtDNA and rapidly evolving nuclear DNA (micro-and/or minisatellite loci) in Atlantic bluefin tuna should be encouraged. Conclusion There is less genetic information available for Atlantic bluefin tuna than for other scombroid fish. The studies of Edmunds and Sammons (1971, 1973) are consistent with the hypothesis that eastern and western management units of Atlantic bluefin tuna comprise a unit Mendelian population (i.e., they are genetically homogeneous). The remaining studies are either incomplete or inadequate to address the issue. Recommendation A major research effort should be undertaken to thoroughly assess the genetic basis of the population structure of Atlantic bluefin tuna. Multiple genetic characters, detected by a variety of approaches, should be employed to provide information on several fundamental questions and to resolve the issue of stock structure. It is critical to support a variety of genetic studies.

OCR for page 9
An Assessment of Atlantic Bluefin Tuna LIFE HISTORY PARAMETERS This section focuses on aspects of life history of Atlantic bluefin tuna that may be relevant to management. All known aspects of bluefin tuna life history have been discussed in detail by Clay (1990) and will not be reviewed here. The committee holds the view that there are aspects of life history that may influence catch per unit effort (CPUE) or other indices used in stock assessment models. We also argue that there are important aspects of life history that are not considered in ICCAT's Standing Committee on Research and Statistics (SCRS) deliberations. Further, much of the life history of Atlantic bluefin tuna is not known. Life history variables such as age composition, growth, age at maturity, and mortality have been used to infer the population structure of several fish (Ihssen et al., 1981). When used as evidence for two populations, these measures are indirect indicators of possible genetic differences. They also can reflect individual responses to environmental differences among localities, so that conclusions concerning population structure based on these data are suspect since they cannot differentiate between genetic and environmental influences. If differences in life history variables result entirely from environmental factors, the two-population hypothesis cannot be tested with these values. In theory, natural selection, genetic drift, and migration between or among localities determine the degree that life history variables change from one locality to the next: high levels of migration tend to minimize differences among localities, whereas extremely low levels of migration could allow population differences to appear in only a few generations. Geographic Locality of Spawning Grounds Although bluefin tuna have been found as far north as Newfoundland in the western Atlantic Ocean and as far north as Norway in the eastern Atlantic Ocean, and a fishery existed for a short time as far south as Brazil, extensive searching has detected only two spawning localities: the Gulf of Mexico and the Mediterranean Sea (Figure 2-1). Each of these localities is large relative to the spawning areas of many other fish species, but small relative to the spawning areas of tropical tunas. Individual females in both the east and the west produce about 30,000,000 eggs each (Clay, 1990). There is no evidence that the large geographic separation of the spawning localities represents reproductive separation. Richards has reviewed evidence of spawning in the Gulf of Mexico (Richards, 1976) and summarized results of ichthyoplankton surveys in the western Atlantic Ocean (Richards, 1987). Larvae and juveniles are found primarily in the northern region of the Gulf of Mexico, with sporadic occurrences in the Florida Straits and off the Texas coast. Larvae have been sampled off the Carolina coast in the western Atlantic Ocean, but their presence there may result from advection by currents from the Florida Straits and not from local spawning

OCR for page 9
An Assessment of Atlantic Bluefin Tuna (Richards, 1990). No bluefin tuna larvae have been found in the eastern Atlantic Ocean (Cort and Loirzou, 1990a), and it has been assumed that bluefin tuna do not spawn there. The western Mediterranean Sea and the Adriatic Sea appear to constitute the second major spawning location for bluefin tuna, but the highest concentrations of bluefin larvae are found in the central part of the western basin of the Mediterranean Sea between southern Italy and Sardinia and around the Balearic Islands of Spain (Cort and Loirzou, 1990a). There are some key questions regarding these spawning locations. What are the biological requirements for spawning, and what environmental factors trigger spawning? How much spawning occurs in one location relative to another, and does it vary from year to year because of changes in environmental conditions? A scientific effort should be made to learn the biological and environmental requirements of spawning. Some scientific effort should also be directed toward estimating the relative amounts of spawning in the two locations, by using identical survey methods in both locations. Timing of Spawning Most aspects of spawning in Atlantic bluefin tuna are still unknown because spawning has not been observed. It is not known if bluefin tuna spawn once or many times per season or whether an individual spawns yearly. It is also not known whether individuals can spawn in the east and then in the west at different times. Spawning in the Gulf of Mexico reportedly occurs from mid-April to mid-June (Richards, 1990; Dicenta et al., 1980). Spawning in the Mediterranean Sea is thought to occur from June to August (Rodriguez-Roda, 1971; Dicenta et al., 1980; Cort and Loirzou, 1990b). It is not known, however, whether later spawning times for the Mediterranean fish are based on genetic differentiation or whether they are in response to environmental differences between the two locations. Differences in spawning times do not necessarily indicate that bluefin tuna produced in the Gulf of Mexico and maturing there or in the western Atlantic Ocean would be precluded from spawning, as adults, in the Mediterranean Sea, or vice versa. Tagging experiments demonstrated that fish can cross the Atlantic Ocean in less than 60 days. It is possible for a fish to spawn in the west in April, migrate to the east, and arrive in time to spawn in the east the same year. Earlier studies providing information on water temperature during spawning of bluefin tuna are referred to in Tiews' (1963) review of biological data on bluefin tuna. A range of 24.9°C to 29.5°C is reported for the Straits of Florida, from Havana to Bimini (Rivas 1954). In the central Mediterranean Sea, the reported range is 19°C to 21.6°C (Roule 1924). Also, for the Mediterranean Sea, large and small sized, sexually mature bluefin tuna spawn at different water

OCR for page 9
An Assessment of Atlantic Bluefin Tuna Depletion in any particular location reflects large-scale patchiness (in time and space) of bluefin tuna abundance and is related to environmental shifts or prey species also responding to environmental changes or both. One of the columns in Table 2-2 (column 6, P.E.I. Gulf) is used as an index in the VPA model. Given either of the above interpretations of the data, the inclusion of the Prince Edward Island (P.E.I.) tended-line data in the VPA model seems tenuous, because the decline of abundance in this area may not represent abundances at other locations. Giant Bluefin Tuna in Winter Little is known about movements after the autumn feeding along the eastern coast of the United States and Canada. The success of a longline fishery in the mid-Atlantic Ocean region suggests movement from the northwest Atlantic Ocean to the mid-Atlantic Ocean region, or from the northeast Atlantic Ocean to the mid-Atlantic Ocean region. During the U.S. exploratory fishing trials from 1957 to 1965, giant bluefin tuna were caught in the mid-Atlantic Ocean region in May but were not caught in January and February (Wilson and Bartlett, 1967). The mid-Atlantic fishery apparently catches fish that moved into this area since the 1960s, perhaps related to changes in the ocean environment. Movements of Age 0 Fish, Small Fish, and Medium Fish6 These movements are driven by searches for water with favorable temperature and food rather than by reproduction. Age 0 fish move from the Gulf of Mexico as far north as Cape Cod. "Small" bluefin tuna migrate slightly farther north. Some small fish apparently move as far east as the eastern Atlantic Ocean; that is, they mix with those spawned in the east. From the eastern spawning ground, small fish move northward to the Bay of Biscay and southward to the Canary Islands. Some move as far west as the western Atlantic Ocean; that is, they mix with those spawned in the west. Little is known about the movements of "medium" fish, except that they move farther eastward offshore than small fish. Practically nothing is known about movements in winter. Presumably fish move toward shore and up the coast following food or particular water temperatures. The limited evidence suggests that the distributions of small and medium fish also are patchy with scales of patchiness being large in time and space. 6    Age 0 fish are described as less than 50cm and less than 3kg, small fish are 50 to 129cm and 3 to 44kg, and medium fish are 130 to 180cm and 45 to 130kg (Clay, 1990).

OCR for page 9
An Assessment of Atlantic Bluefin Tuna Giant Bluefin Tuna in New England Recent observations by New England commercial and recreational fishermen of large increases in local abundance may reflect large-scale changes in a food patch or substock. The New England patch may diminish within 10 years regardless of any regulations that are imposed on the fishery. The change may not represent the depletion of the stock, but rather the depletion of a food patch or other environmental changes that result in the redistribution of giant tuna. Conclusions There are several aspects of life history that influence the abundance indices used in management, and many aspects of life history are unknown or poorly known. There is no evidence that the large geographic separation of spawning locations represents reproductively isolated spawning grounds. Giant bluefin tuna tend to occur in patches where food is abundant and in areas of suitable temperatures. The locations of these patches appear to change from year to year. Recommendations NMFS, in cooperation with ICCAT, should make an effort to understand the biological and environmental requirements of spawning. Some effort should be made to estimate the relative amounts of spawning in the two locations, using identical survey methods. Additional studies of age and size at spawning should be carried out. In particular, the committee suggests that all fish caught as by-catch in the Gulf of Mexico be sampled: the gonads should be taken for histological analysis, and body length should be measured to estimate age. Because bluefin tuna are part of an ecosystem, the committee recommends the use of the ecosystem approach to management. For example, some effort should be directed toward estimating major changes in distribution and abundance of prey species, as prey abundance seems likely to alter the distribution and abundance of bluefin tuna. This recommendation applies to all the life history stages of bluefin tuna. This information would not necessarily be used as an index but could be useful in choosing or weighting indices or in designing various sampling schemes. A tagging program should be undertaken by NMFS, in cooperation with ICCAT to provide better estimates of the magnitude and patterns of movement (refer to the set of design features discussed in Appendix D). This program should be designed to answer specific scientific questions pertinent to stock assessment. The program should be coordinated among all nations participating in ICCAT studies. Tagging should include appropriate combinations of conventional, PIT, acoustic, and archival tags (see Appendix D for a description of

OCR for page 9
An Assessment of Atlantic Bluefin Tuna archival tags). Archival tags will provide information regarding fidelity to particular spawning locations and/or particular feeding locations (i.e., fishing grounds). A well-designed program using conventional tags will provide better information regarding mixing between the eastern and western Atlantic Ocean as well as better information regarding the extent that mixing changes with fish age and how it varies from year to year. It has been suggested that an aerial survey could be developed and used as a fisheries-independent estimate of abundance. Acoustic tags may be useful in calibrating aerial survey data by providing information about the vertical distribution of bluefin tuna. PIT tags (passively Integrated Transponder) or coded wire tags may be useful in a forensic sense to estimate the magnitude of nonreporting of tag recaptures. The committee emphasizes the importance of planning tagging experiments. Much of the available information suggests that the present tagging data resulted from opportunistic programs. The committee suggests a pooling of resources among participating nations to effect a strong tagging program. Changes in tuna distribution over time could be explained, at least in part, by large-scale changes (in time and space) in environmental conditions. These changes are discussed in the next section. CLIMATE Climate and Evolutionary History One assumption underlying genetic methods used to resolve the population structure of bluefin tuna is that any of the populations are in equilibrium with migration and genetic drift (random processes). However, equilibrium is achieved, on average, only after N generations (Kimura, 1955), which is on the order of millions of years for marine fish. This assumption of equilibrium is probably not completely met for North Atlantic pelagic fish, in large part because the present-day population structure in the North Atlantic Ocean has arisen since the last glaciation (15,000 to 20,000 years ago), when an ice sheet covered the North Atlantic Ocean as far south as New England in the western Atlantic Ocean and the British Isles in the eastern Atlantic Ocean (CLIMAP, 1976). At that time, temperatures in Caribbean waters were about 5°C cooler than at present (Guilderson et al., 1994), and the present spawning areas of bluefin tuna most likely were displaced geographically. Gene flow and the past glacial history of the North Atlantic Ocean may have been important in homogenizing populations of bluefin tuna in the North Atlantic Ocean through population contractions and recolonizations. As populations of marine fish decline, their geographic ranges contract (MacCall 1988), and gene flow over the smaller area homogenizes any genetic differences. When oceanographic conditions become more favorable, genetically homogeneous stocks recolonize the basin. Genetic differentiation has not been observed in

OCR for page 9
An Assessment of Atlantic Bluefin Tuna Atlantic herring (Clupea harengus [Grant, 1984]) nor in Atlantic cod (Gadus morhua [Grant and Stahl, 1988]), even though it is unlikely that transocean movement presently occurs in these fish. Thus, the lack of allozyme differentiation in Atlantic bluefin tuna may reflect the recent glacial history of the North Atlantic Ocean rather than ongoing migration across the Atlantic Ocean. Even if no migration occurred between western and eastern populations, allozyme frequency differences may not appear for several thousand generations. Mitochondrial and microsatellite DNA analyses, which are theoretically capable of detecting population structure on a finer scale than allozyme or mtDNA analyses, may be required to resolve the component of genetic population structure of bluefin tuna resulting from recent population events. Conclusion Past climatic events can have long-lasting effects on the population genetic structures of present-day populations. Molecular techniques used in studies of population structure must be capable of distinguishing historical versus contemporary factors. Fish Abundance and Climatic Changes Climatic variability on a shorter time scale can also influence bluefin tuna populations, but separating the effects of fishing from those of climatic change continues to challenge fisheries scientists. Although the regional abundances of some fish appear to be linked to oceanographic changes, underlying mechanisms controlling abundance are difficult to understand with data from studies that are geographically limited or of short duration. In the Pacific Ocean, fisheries fluctuations, in association with El Niño Southern Oscillations, have long been recognized. Abundances of anchoveta, and associated higher trophic levels, periodically increase in response to nutrient enrichment associated with upwelling (Barber and Smith, 1981). These upwelling events in turn, are apparently responding to pressure-driven oscillations, on the order of three to seven years, in the sea level of the western equatorial Pacific (Wyrtki, 1977). On an interdecadal scale, the abundances of sardines and anchovies in many areas are cyclic, and are often inversely related to each other where they co-occur either in recent times (Parrish et al., 1983) or historically (Shackleton, 1987). Historical fishing records of the Japanese sardine fishery dating to the 1600s show marked fluctuations in sardine abundance (Cushing and Dickson, 1976). Strong fluctuations in abundance over the past 2,000 years have also been inferred from variability in sardine scale abundance in anaerobic sediments off California (Soutar and Isaacs, 1974; Baumgartner et al., 1992), Peru (De Vries and Pearcy, 1982), and southern Africa (Shackleton, 1987). These fluctuations in abundance predate intensive fishing on these species.

OCR for page 9
An Assessment of Atlantic Bluefin Tuna FIGURE 2-4  Barometric shifts and wind patterns driving the Russell Cycle (from Mann and Lazier [1991], with permission of Blackwell Scientific Publications Ltd., Oxford, England). Another example of the effect of interdecadal climatic change on marine populations is the Russell Cycle in the North Atlantic. Russell observed marked changes in the abundances of macrozooplankton in the English Channel between 1924 and 1971, and suggested that local changes in abundance resulted from distributional changes associated with recurring climatically-driven oceanographic

OCR for page 9
An Assessment of Atlantic Bluefin Tuna changes in the North Atlantic Ocean (Cushing and Dickson, 1976). The North Atlantic Ocean is alternately dominated by high-and low-pressure systems with a periodicity of 10-20 years. When a high-pressure system dominates, the prevailing westerly winds are diverted north of the high, resulting in a northeastward flowing body of warm air and water along North America in the western Atlantic Ocean and a southeastward flowing body of cool air and water along Europe and the British Isles (Figure 2-4). When a low-pressure system dominates in the North Atlantic Ocean, the westerlies are diverted to the south, and this results in warmer weather and sea temperature in the eastern Atlantic Ocean (Figure 2-4). A low-pressure system dominated the North Atlantic Ocean from about 1880-1935 and led to a period of gradual warming in Europe, culminating in the dramatic decade of 1925 to 1935, with the appearances of numerous tropical species along European shores and temperate species in Scandinavian waters. At this time, fishing for coldwater gadid fish was adversely influenced, and the abundance of bluefin tuna increased near Iceland (Cushing and Dickson, 1976) and in Norwegian waters (Tiews, 1978). During this period, geographic ranges of many intertidal invertebrates were displaced to the north in the British Isles (Southward, 1980). This warm phase ended about 1935 and was followed by about 15 years of average or variable conditions, after which the North Atlantic Ocean was dominated by a high pressure system from about 1950-1970. Cooler oceanographic conditions persisted in the eastern Atlantic Ocean and gadid abundances there increased sharply despite heavy fishing. Intertidal invertebrate distributions moved southward, with coldwater boreal species replacing warmwater southern species in the British Isles (Southward, 1980). Since 1970, a low-pressure system has dominated the North Atlantic Ocean. British intertidal invertebrate ranges shifted back to the north and the fishery for gadid fishes has collapsed. The oceanographic conditions in the western North Atlantic Ocean are more complex than those in the eastern North Atlantic Ocean. Periods of cold weather in the eastern North Atlantic Ocean are often associated with periods of warm weather in the western North Atlantic Ocean, but periods of warm eastern Atlantic Ocean weather are not always associated with cold waters in the western North Atlantic Ocean. Even though fish abundances in the western North Atlantic Ocean are not strongly correlated with climatic cycles, anecdotal reports of geographic changes in the distributions of species still abound. Tilefish, bluefish, and menhaden, which are harvested chiefly off the southeastern United States, have periodically been abundant off New England (Cushing, 1982). Hopkins and Garfield (1979), in a long-term study of temperature trends in the Gulf of Maine, suggested cycles of slightly longer than 20 years, with temperature and salinity minima in the early 1940s and the mid 1960s. This region of the western North Atlantic Ocean may be strongly influenced by the cold, southward flowing Labrador Current and by warm core rings of the northeastward-flowing

OCR for page 9
An Assessment of Atlantic Bluefin Tuna Gulf Stream. Eddies of warm Gulf-Stream water can vary in number and duration before being absorbed back into the Gulf Stream (Richardson, 1983). Although tuna are highly migratory, they are affected by climatic and oceanographic changes. Yellowfin, bigeye, and albacore, for example, show stronger year classes when sea surface temperatures are cool (Yamanaka and Yamanaka, 1970, cited in Cushing, 1982). The thermal boundaries of adult yellowfin tuna occur between 18° and 31°C, but commercial concentrations occur between 20° and 28°C (Cole, 1980). Smaller individuals of Atlantic bluefin tuna also appear to be restricted to areas in the Gulf of Maine with surface temperatures of at least 16°C, whereas adults are found in waters as cold as 10-12°C, which appears to be their lower thermal limit in the Gulf of Maine (Bigelow and Schroeder, 1953). Similarly, 12°C also appears to be the lower level of thermal tolerance for bluefin tuna in the eastern North Atlantic Ocean (Luhmann, 1959; Tiews, 1962). Giant fish however, are found at temperatures as low as 6°C off Newfoundland (Squire, 1962). Changes in oceanic temperature distributions have apparently influenced the local abundances and distributions of bluefin tuna. Bluefin tuna were first seen in Norwegian waters about 1907 at the beginning of a warming trend. This fishery peaked in 1952, with the beginning of a cooling trend, and collapsed in 1963. The collapse was attributed to overfishing (Tiews, 1978), but may have been, at least in part, owing to long-term climatic change. Farther to the south, the abundances of bluefin tuna peaked in Portuguese waters in about 1880, then decreased to one tenth this size by 1920 (Neuparth, 1925). Abundance peaked again in 1937, and declined from 1958 to 1963. Although these cycles in abundance are not as strongly correlated with the Russell Cycle as are the Norwegian abundances, the declining trends may not be associated with overfishing. These results suggest two conclusions. First, changes in abundance and geographic distribution are influenced by both climate and fishing. When abundances decline, the geographic distribution of bluefin tuna may shrink to the center of its distributional range, so that local abundance may not decline in some areas even though the overall stock is declining. Second, even though the distributions of bluefin tuna respond to changes in sea surface temperatures, attempting to attribute changes in abundance solely to interdecadal changes in temperature may be an oversimplification. Cool water temperatures may also result from changes in upwelling, which also brings nutrient rich water to the surface and may enhance the productivity of plankton, which, in turn, may influence tuna larval abundance and year-class strengths. Climatic changes may also increase terrestrial precipitation and associated runoff with micronutrients, such as iron, that limit primary production (Gran, 1931; Martin, 1992). The important point is that climatic changes occur over wide areas and periods of many years, and these changes can affect the distributions

OCR for page 9
An Assessment of Atlantic Bluefin Tuna and abundances of fish species. Fisheries managers must recognize that climatic changes can modify, for better or worse, the adverse influence of heavy fishing. Conclusion Changes in distribution and abundance and local declines in abundance can be brought about by changes in ocean climate, which may confound the effects of exploitation. Recommendation Adaptive management strategies that account for the effects of both climatic changes and exploitation should be considered for implementation. Such strategies should also account for the dynamic nature of geographic distributions in space and time. MOVEMENT BASED ON NONGENETIC MARKERS Nongenetic markers can provide only indirect evidence for gene flow across the Atlantic Ocean because there is no assurance that fish moving from one area to the next actually breed in the new area; mixing on the fishing grounds may not necessarily reflect mixing of the same magnitude, or mixing of the same fish, on the spawning grounds. Such data can, nonetheless, provide circumstantial evidence for potential interbreeding among stocks. Naturally occurring markers of population mixing include parasites and chemical constituents, both of which must be acquired at an early age in nursery areas to be useful for discrimination among stocks. Tag and recapture experiments are specific attempts to estimate movement from one area to another and can be used at any age of the fish. Such data constitute conclusive evidence of movement from the area of tagging to the area of recapture. Parasite Markers The geographic distributions of parasites on marine fish have been used to infer the evolution and migratory patterns of the fish (e.g., Kabata and Ho, 1981). For this approach to be useful in determining the geographic origin of a fish, the fish must acquire a parasite at an early age and the parasite must have a restricted geographic distribution where infection is possible. The trematode Nasicola sp. infests the nasal pores of the bluefin tuna and is tropical, so that infection is possible only in the Gulf of Mexico or in other tropical waters of the Atlantic Ocean. The copepod Elytrophora sp. infests the gill chamber of bluefin tuna, lives in temperate waters, and cannot be acquired in the tropics. Walters (1980) conducted a study of these parasites in Atlantic bluefin tuna and the results of his

OCR for page 9
An Assessment of Atlantic Bluefin Tuna work are often cited as evidence of movement of bluefin tuna between areas in the north Atlantic Ocean. A close scrutiny of these data, however, show that his results cannot be used to infer movement (Appendix E). Conclusion Existing data on parasite markers do not provide evidence to support the mixing hypothesis. Recommendation This avenue of research is not recommended until experts in the field can show its utility and efficacy for estimating movement of bluefin tuna. Microconstituent Analysis The idea behind this method of estimating population structure is that the proportions of chemical elements incorporated into fish bones differ from one area to the next because of differences in elemental compositions of prey and seawater. Thus, each nursery area may have a different elemental signature in bony material laid down during growth, and the signature can be used to identify the origins of fish captured outside the nursery areas. Calaprice (1985) and Calaprice et al. (1971) measured elemental signatures by x-ray fluorospectroscopy. Energy from gamma rays is absorbed by electrons in the various elements embedded in bone and is spontaneously released to produce characteristic x-ray signatures at lower energies. The x-ray emission spectrum is recorded and used to characterize individual fish. The two Calaprice studies are concordant with the hypothesis of significant movement between east and west. However, they cannot be used to assign a value to the magnitude of the mixing because the results vary markedly between studies. The reason for the variability of the studies probably resides in the nonrandom procedures used to obtain fish samples from the different fishing grounds. The rationale for the committee's position is detailed in Appendix F. Conclusions The microconstituent studies support the hypothesis of mixing between the two regions. Some giant fish harvested in the western Atlantic Ocean were spawned in the Mediterranean Sea, and some that were harvested in the eastern Atlantic Ocean were spawned in the Gulf of Mexico. The data suggest that the birthplace of older tuna can be estimated by analyzing the center of the vertebrae because some aspects of the chemical ''fingerprint" are stable over time.

OCR for page 9
An Assessment of Atlantic Bluefin Tuna The available data and analysis cannot be used to ascribe a value to the magnitude of mixing in either direction. Recommendation This avenue of research should be pursued by NMFS and ICCAT because the preliminary results look promising. There have been marked improvements in techniques similar to those employed by Calaprice and in software used to discern patterns. Any research proposal regarding this complex analytical technique should be peer reviewed. Calaprice noted that "the acquisition of adequate samples has been a difficult and limiting task." Any future research should engage all nations so that adequate samples from all areas can be used in the analysis. A single-blind approach should be used so that the persons doing the analysis do not know the source of the material but are aware of the number of localities that have been sampled. The observation that bluefin tuna often are found in schools (or larger patches) means that each sample must cut across schools and patches. A random sample from the western Atlantic Ocean must include samples from several localities, and not just from a single school.

OCR for page 9
An Assessment of Atlantic Bluefin Tuna This page in the original is blank.