A REVIEW OF THE BIOLOGY AND FISHERIES FOR YELLOWFIN TUNA, THUNNUS ALBACARES, IN THE EASTERN PACIFIC OCEAN

Alex Wild
Inter-American Tropical Tuna Commission
La Jolla, California

1. INTRODUCTION

This paper is a review of the biology, resource and fisheries associated with yellowfin tuna, Thunnus albacares, in the eastern Pacific Ocean. Under certain headings, such as Early Life History and Age and Growth, information from the central and western Pacific and the eastern Atlantic has also been included to offer a wider perspective. The material in the section on Gear Interaction is intended to be both a general review and an interaction-specific review for the purse-seine and longline gears operating in the eastern Pacific. The task of compiling the material was simplified considerably through the extensive use of two major sources of information: the synopsis of biological data on yellowfin by Cole (1980), and the many anonymous contributions made to the Inter-American Tropical Tuna Commission's (IATTC's) Annual Reports.

2. CLASSIFICATION

Among fisheries biologists it is generally accepted that the expression “tunas and related species” applies to all genera of the family Scombridae, except Rastrelliger and Scomber, and to all genera of the families Istiophoridae and Xiphiidae (Klawe, 1980). The mackerels, Rastrelliger and Scomber, are excluded because of their mode of life. The nature of the fisheries directed toward mackerels, and the marketing of their catches, differ considerably from those of the tunas and related species. The listing below identifies the position of yellowfin tuna within the family Scombridae, following Lindberg's (1971) scheme to the family level:

Phylum Chordata
    Subphylum Vertebrata
        Superclass Gnathostomata
            Class Teleostomi
                Subclass Actinopterygii
                    Order Perciformes
                        Suborder Scombroidei
                            Family Scombridae
                                Sub-family Scombrinae
                                    Tribe Thunnini
                                        Genus Thunnus
                                            Species albacares

The classification of yellowfin and other tunas is discussed by Gibbs and Collette (1967), Sharp and Pirages (1978), and Collette and Nauen (1983).

3. EARLY LIFE HISTORY

The distribution of yellowfin eggs has not been determined because it is not possible to differentiate them from those of many other scombrids (Cole, 1980).

The development from the pre-larval to the post-larval phase of yellowfin hatched from artificially fertilized eggs has been described by Harada et al. (1971), Mori et al. (1971) and Harada et al. (1980). Descriptions of larval and post-larval yellowfin collected at sea are given by Mead (1951), Wade (1951), Matsumoto (1958, 1962), Sun' (1960), Ueyanagi (1966, 1969) and Matsumoto et al. (1972). The internal anatomy of yellowfin larvae has been described by Richards and Dove (1971). The juvenile forms have been described by Schaefer and Marr (1948), Wade (1950a), Yabe et al. (1958) and Matsumoto (1961). Guides to the identification of early life history stages are offered by Fahay (1983), Collette et al. (1984), Nishikawa and Rimmer (1987) and Richards (1989). Yellowfin larvae in the total length (TL) range of 3–10 mm can be distinguished from those of bigeye tuna (T. obesus) by the presence in bigeye of a single melanophore on the ventral tail (Richards et al., 1990). The distinction becomes obscured during transformation from larvae to juveniles because both species may then display the melanophore. Graves et al. (1988), however, successfully applied an electrophoretic technique to an apparently mixed-species sample and identified all of the small (10–12 mm) juveniles as yellowfin.

Yellowfin larvae distribution has been examined by Wade (1951), Sun' (1960), Yabe and Ueyanagi (1962), Ueyanagi et al. (1969) and Mori (1970) for the western Pacific; by Matsumoto (1958), Strasburg (1960), Sun' (1960) and Nakamura and Matsumoto (1967) for the central Pacific; and by Mead (1951), Klawe (1963) and Klawe et al. (1970) for the eastern Pacific. Much of the information given by these investigators has been incorporated into the works of Yabe et al. (1963), Matsumoto (1966), Ueyanagi (1969), Nishikawa et al. (1985) and Suzuki et al. (1978) in investigations of the distribution of yellowfin larvae in the entire Pacific Ocean. These studies indicate that the larvae are trans-Pacific in occurrence, although their distribution is limited latitudinally to tropical and subtropical waters. Larvae occur year-round in equatorial waters, but there is a seasonal change in density in the subtropical waters of the central and western Pacific. Seasonal peaks in density of larvae occur in the Kuroshio Current area during May to June and in the East Australian Current during November to December. In the eastern Pacific the range of yellowfin larvae is compressed somewhat during the northern winter by cold water converging toward the equator from the north and south. Although data are not available for an entire year, on a quarterly basis there appears to be a peak in larval density from April to June off Central America (Figure 1).

Klawe (1963) examined the vertical distribution of yellowfin larvae in the eastern Pacific. He found no evidence of occurrence of larvae below the thermocline. Information on the vertical distribution of yellowfin larvae in the central Pacific has been provided by Matsumoto (1958) and Strasburg (1960), and in the Indo-Pacific Ocean by Ueyanagi (1969). These investigators indicated that yellowfin larvae are probably restricted to the upper 50 or 60 metres of the ocean. Richards and Simmons (1971) found that in the northwestern Gulf of Guinea and off Sierra Leone, yellowfin and bigeye tuna larvae migrate to the surface during the day, while the larvae of skipjack tuna (Katsuwonus pelamis) migrate to the surface at night. As an aside, Davis et al. (1990) reported that skipjack larvae behave similarly in the east Indian Ocean, whereas southern bluefin (T. maccoyii) and albacore (T. alalunga) larvae migrate to the surface layer in the daytime.

Figure 1. Quarterly density distributions of larval yellowfin sampled by surface horizontal tows. Solid and open circles denote the densities in 5-degree areas calculated from five or more tows and less than five tows respectively. N indicates the number of nominal tows and the numerals represent areas (shown surrounded by heavy lines) for examining seasonal changes of density in them. (Reproduced from Suzuki et al., 1978).

Higgins (1967) summarized the published accounts on the capture of juvenile yellowfin. He reported that they have been collected in the western Pacific as far north as approximately 31°N, near the coast of Japan, and as far south as 23°S. In the central Pacific the northernmost record is 23°N, near the Hawaiian Islands, and the southernmost record is 23°S. In the eastern Pacific juvenile yellowfin have been recorded from approximately 24°N, off Baja California, to approximately 2°S, off the coast of Ecuador. Higgins (1967) found no records of juvenile yellowfin in the area bounded by 150°W and 112°W, but he attributed this hiatus to insufficient sampling in this area. Higgins (1970) also examined the distribution of juvenile tunas in Hawaiian waters from July to September, 1967. Although skipjack was the dominant species encountered, he found that juvenile yellowfin were more abundant offshore than inshore, and that they exhibited no extensive vertical migrations.

4. AGE AND GROWTH

4.1 Age Determination and Growth

The history of age determination of Pacific tunas has been summarized by Hayashi (1957), Bell (1964), Shomura (1966) and Suzuki (1971). The benefits and limitations associated with the scale-reading method were reviewed by Suzuki (1974). At first, the principal methods of estimating age relied on the growth marks in hard parts, such as vertebrae (Aikawa and Kato, 1938) and scales (Nose et al., 1957; Yabuta et al., 1960; Yang et al., 1969), but a consistent and accurate interpretation of age was hampered by the lack of an independent method of verification. Emphasis was therefore directed to the assessment of growth by means of size-frequency modal analysis.

In the eastern Pacific (Hennemuth, 1961a; Davidoff, 1963; Diaz, 1963), eastern Atlantic (Le Guen and Sakagawa, 1973) and western Pacific (Wankowski, 1981) the analyses were based on fork-length frequencies, while Moore (1951) in the central Pacific utilized weight frequencies. The resulting growth curves were affected to some degree by the uncertainties generated by spawning periods of extensive duration, and by the need to anchor the growth curves to the time axis by a suitable choice of age for a particular size. In the eastern Pacific the results were also influenced by the relative lack of large fish (> 100 cm fork length) in the catches during the late 1950s and early 1960s, and by the suspicion of gear selectivity for sizes < 70 cm. Despite these difficulties, there was a gradual convergence of size-at-age estimates prepared by the scale and modal progression methods (Suzuki, 1971).

Mark-recapture experiments involving tetracycline injection at sea demonstrated that eastern Pacific yellowfin north of the equator deposit growth increments daily on their otoliths in the thawed fork-length range from 40 to 110 cm (Wild and Foreman, 1980). Based on this result, the following growth equation was developed for combined sexes over the length (l_t) range of 30 to 168 cm,

by estimating the ages in days of 196 yellowfin sampled from 1977 through 1979 (Wild, 1986). To avoid the effects of size selection and differential mortality, each 10-cm interval of the growth curve contained about 15 items. Additional tetracycline, mark-recapture experiments have since extended the validated range of daily deposition to 148 cm (IATTC, 1988) and confirmed about 65 percent of the growth curve from 0–180 cm, i.e. from 30–148 cm. Circumstantial evidence suggests that the ratio of increments to days is also 1:1 in the range from 0 to 30 cm and from 148 to 168 cm, and the estimates of ages are therefore believed to be accurate. In the above equation, weighted regressions identified the Richards (1959) function as providing the best statistical fit to the data from the following choice of models:

where ι_t is the fork length at time t; L∞ is the asymptotic size; K is a growth parameter; t_o is the time-axis intercept; t_* is the inflection point of the curve; and m is a shape. parameter related to the ratio of ι_t/L∞ at t_*. Either the Gompertz or the Richards model provided a significantly improved fit to the growth data of Moore (1951), Hennemuth (1961a), Davidoff (1963) and Le Guen and Sakagawa (1973), once the constraint of the non-inflective von Bertalanffy model was removed. The comparative growth curves shown in Figure 2 include these adjustments. The growth equation parameters and estimates of sizes at different ages appear in Table 1. Counts of otolith increments were also used by Uchiyama and Struhsaker (1981) to estimate the ages of 14 central Pacific yellowfin. Over the length range from 6 to 93 cm, growth was best described by two linear stanzas. Draganik and Pelczarski (1984) also estimated the growth of central Atlantic yellowfin by assuming that the annuli in sectioned dorsal fin rays were equivalent to years.

Sexual dimorphism has been detected in eastern Pacific yellowfin in terms of growth in length, weight and the length of the counting path on the otolith (Wild, 1986). After about 95 cm in length, the growth rate of females becomes progressively slower than that of males. Although these sexual differences are statistically significant, they are small and may be unimportant in a practical sense. Growth for individual sexes was best described by the Gompertz model, and the equation parameters for length appear in Table 1, together with Le Guen and Sakagawa's (1973) estimates for western Pacific males and females calculated from the data of Yabuta et al. (1960). For combined sexes the annual variation in growth in length was significant (Wild, 1986), and confirmed earlier reports of annual variation by Hennemuth (1961a), Davidoff (1963), and Le Guen and Sakagawa (1973). In the length range from 30 to 110 cm, yellowfin caught inshore were significantly heavier than those caught offshore, but the situation was reversed for fish longer than 110 cm (Wild, 1986). This result suggests that there may be size-related changes in the diet of yellowfin that are associated with feeding opportunities in the two areas. In the same study, the following equations describe growth in weight (w) for combined sexes of thawed, whole fish;

and the relationship between length and weight for combined sexes in the inshore-offshore regions;

Although this last equation is applicable over a broad length range, it is based only on 196 fish. In addition, the weights of small fish (30–40 cm) tend to be overestimated somewhat due to a slight, but insignificant, degree of nonlinearity. On the other hand, the equation developed by Chatwin (1959) for pooled data in the eastern Pacific,

is applicable over a much narrower length range, and it is also weighted toward the dominant sizes of yellowfin (50–85 cm) available to the baitboat fishery during 1956–57. Although significant differences existed among relationships for different areas, they were not considered biologically important at the time. After a conversion to natural logarithms, the length-weight equation reported by Nakamura and Uchiyama (1966) for central Pacific yellowfin, i.e.,

underestimates the weight in the eastern Pacific for a given length. Therefore the relationships for the two regions are probably distinct.

Figure 2. The fork-length growth curve (P) of yellowfin tuna in the eastern Pacific based on otolith-increment age estimates compared to previous curves derived from size-frequency analysis. (Reproduced from Wild, 1986.)

TABLE 1. Comparison of growth equation parameters and estimated sizes at age for yellowfin tuna for different oceanic regions and by different investigators. Legend: Eq. = equation (see section on Age and Growth in text for growth models); R = Richards (1959); G = Gompertz (Ricker, 1979); B = von Bertalanffy (1938); L = linear segments; f = fixed parameter. Bracketed quantities represent extrapolations slightly beyond the range of the data.

4.2 Growth Rates

The growth rates of yellowfin in the eastern Pacific have also been estimated by means of tag-recapture experiments conducted from 1955 through 1981 (Bayliff, 1988). The rates for particular length groups appear to have annual and areal components, but they are not persistent, and it is therefore more informative to mention general trends. Overall, the growth rate (± standard error) of tagged yellowfin in the length range from 25 to 100 cm is about 0.85 ± .01 mm/day. Growth appears to be most rapid near the Revillagigedo Islands (1.11 ± .03 mm/day) and slower off Baja California (0.69 ± .02 mm/day), Central America-Colombia (0.69 ± .02 mm/day) and Ecuador-Peru (0.52 ± .10 mm/day). Over the length range from 25 to 100 cm, growth for tagged fish was nearly linear and parameters for curvilinear models could not be calculated. In contrast to growth rates estimated from modal analysis for small yellowfin in the eastern Atlantic (40–70 cm: Fonteneau, 1980; 35–65 cm: Bard, 1984), the southwestern Pacific (30–50 cm; Brouard et al., 1984) and the western Indian Ocean (Marsac and Lablanche, 1985), juvenile yellowfin in the eastern Pacific do not exhibit a period of slow or reduced growth rate, i.e. ≤ 0.5 mm/day. By means of otolith increments Yamanaka (1990) also demonstrated that yellowfin near the southern Philippines grow rapidly, 2.5 mm/day, in the length range of 15–55 cm, and 0.96 mm/day in the interval from 55–79 cm.

To facilitate comparisons of growth rates obtained from tagging, otolith increment counts and modal progressions, the equation (1) describing growth of eastern Pacific yellowfin for combined sexes (Wild, 1986) was differentiated to yield instantaneous rates and plotted versus age and length (Figure 3). In general, for fish caught north of the equator, the average rate for the untagged yellowfin used in the otolith study was ≥ 1 mm/day, or ≥ 3.04 cm/mo., during growth in the length range from 50 to 115 cm. This rate compares favourably with the average value of 1.00 mm/day derived from the measured change in otolith sizes of tagged and tetracycline-injected fish (Wild and Foreman, 1980). Bayliff (1988) reports that for yellowfin ≥ 50 cm tagged north of the equator, the average growth rate was 0.85 mm/day based on the change in fork length.

5. MATURATION, SPAWNING AND SEX RATIO

5.1 Maturation

Wade (1950b) found a ripe male 525 mm in length near the Philippine Islands. Buñag (1956) encountered a mature female of a similar size (567 mm) in the same area, and also described seven stages of maturation, including the resting stage, based on the diameter of ova preserved in formalin. On the other hand, by means of gonado-somatic indices, Yuen and June (1957) found that in the central equatorial Pacific a few yellowfin reach maturity at about 70–80 cm, but the majority do not mature until they reach 120 cm. Similarly, Kikawa (1962) reported that female yellowfin attain first maturity at more than 110 cm in the western and central Pacific longline grounds, although a few individuals were found to be mature at lengths between 80 to 110 cm. In the eastern Pacific longline grounds east of 130°W, Shingu et al. (1974) found the minimum size at first spawning for both sexes to be between 91 and 100 cm; however, their data also indicated that a high percentage of fish reached first maturity at lengths above 120 cm. Orange (1961) found the minimum size of mature yellowfin along the coast of Central America to be 50 cm; in this area 20 percent of the females in the 50–60 cm size class were found to be mature. More recently, however, the length at sexual maturity in two regions of the eastern Pacific was assessed using histological criteria, the mean diameter of oöcytes in the most advanced mode, and microscopic detection of residual hyaline oöcytes (IATTC, 1990). In the areas bound by 20°–30°N, 110°–120°W, and by 0°–10°N, 80°–90°W, the smallest female found with mature ovaries was 84 cm, and the estimated length at 50-percent maturity was 95 cm.

Figure 3. The instantaneous growth rate of yellowfin tuna in the eastern Pacific relative to the fork length or estimated age of the fish, derived from the growth curve based on otolith-increment age estimates by Wild (1986). The solid curve is used to estimate the instantaneous growth rate from the estimated-age axis, and the dashed curve is used to estimate the growth rate from the fork-length axis.

Kikawa (1966) postulated that estimates of the spawning potential of small (< 100 cm) yellowfin derived from longline-caught fish in open areas of the western Pacific were likely to be low. Suzuki et al. (1978) pointed out that such may also be the case in the eastern Pacific. Significantly higher percentages of sexually mature yellowfin under 120 cm in length were found in samples taken from purse-seine catches than in samples gathered from longline catches made in the same area and month. Hisada (1973) also reported differences in sexual maturity of yellowfin sampled from surface hand-line and longline catches in the Coral Sea. These results seem to imply that each gear type by itself may be an inappropriate sampling device for maturation studies. It is possible that a re-evaluation of the spawning potential of smaller size groups, as suggested by Kikawa (1966) and Suzuki et al. (1978), will show that a greater proportion of fish under 120 cm in open waters of the Pacific are mature than was previously thought.

Kikawa (1966) calculated the following equations for the relationships between body length and ovary weight (both ovaries combined) for yellowfin in the tropical central and western Pacific:

< 160 cm Y = (9.596 × 10-5)X^3.436
> 160 cm Y = 2760 + 5.318X

where Y is ovary weight (gm), and X is fork length (cm). Fecundity studies have been carried out by June (1953) and Joseph (1963) for yellowfin in the central and eastern Pacific, respectively. In the latter area, the relationship between the length and the number of maturing ova is expressed by:

Y = (8.955 × 10-9)X^2.791

where Y is millions of ova in the most advanced mode, and X is the fork length (mm). For example, yellowfin that were 1,000, 1,300 or 1,500 mm long would produce approximately 2.1, 4.4, or 6.6 × 10⁶ ova, respectively.

With the exception of Buñag's (1956) work cited above, the gonado-somatic index, or the ovary weight expressed as a function or percentage of body weight, has been used as the primary indicator of the state of reproductive activity in yellowfin. This index is biased, however, because larger females develop larger ovaries in proportion to body weight than smaller females (de Vlaming, 1982). In his analysis of black skipjack's (Euthynnus lineatus) reproductive biology, Schaefer (1987) stated that if the gonadal index was to be interpreted correctly with respect to reproductive activity, it should be validated with histology and/or oöcyte diameter and adjusted for the size of individuals. Such work is in progress by the IATTC regarding yellowfin in the eastern Pacific.

5.2 Spawning

Ueyanagi (1969, 1978) stated that 26°C is probably the lower limit for spawning of yellowfin. Spawning in the western and central regions of the Pacific Ocean takes place in northern latitudes during the spring and summer of the Northern Hemisphere (Table 2). Spawning takes place year-round in the northern equatorial waters of the western and central Pacific. In southern equatorial waters, however, yellowfin spawn mostly during the first half of the year, with minimal spawning during the second half because of the intrusion of water cooler than 26°C into the area. Off the coasts of Mexico and Central America yellowfin spawn throughout the year, but the peaks of spawning occur at different times in different areas. Spawning appears to be more sporadic and shorter in duration in the coastal spawning areas than in northern equatorial waters.

By identifying the recruitment cohorts present on the spawning grounds, Knudsen (1977) was able to show that yellowfin spawn at least twice a year off the coast of southern Mexico and Central America. Although samples were not available for all months, further offshore the spawning period is at least seven months and may occur all year. The periods of spawning vary in length and time of occurrence from year to year. In preparation for a more intensive study on yellowfin reproduction in the eastern Pacific, the details of which appear in IATTC (1989), Schaefer (1988) estimated that spawning near Clipperton Island occurs between 2000–2400 h. The interval between spawnings was about 1.27 days in samples collected in 1986 and 1.25 days in the 1987 samples.

5.3 Sex Ratio

Everett and Punsly (IATTC; pers. commun.) re-examined sex-ratio data collected in the eastern Pacific from the longline fishery during 1958–62, and from the surface fishery in 1953–62 and 1970–73. Departures from the expected ratio of 1:1 occurred as a result of year, area, gear and length-class effects, but on a quarterly basis within areas and years the data were homogeneous. Despite the complexity of the results, the rapid decline in the percentage of females around 140 cm was a consistent feature of all data sets. Possible explanations for this result are that females may experience differential mortality, growth and vulnerability to the fishery. During development of the sexually-dimorphic growth curves, Wild (1986) did not detect an accumulation of large females in the length range from 130–140 cm, a situation that would result from a cessation of female growth. He therefore attributed the virtual disappearance of females beyond 140 cm to an increase in their natural mortality rate.

TABLE 2. Spawning seasons of yellowfin in the Pacific Ocean. (Reproduced from Cole, 1980).

6. STOCK STRUCTURE, DISTRIBUTION, AND MIGRATION

6.1 Subpopulations and Stocks

Most of the inferences about the subpopulation structure, or self-sustaining genetic units, of yellowfin in the Pacific Ocean have come from indirect sources such as morphometric comparisons (Godsil, 1948; Godsil and Greenhood, 1951; Schaefer, 1952, 1955; Royce, 1953, 1964; Kurogane and Hiyama, 1957; Broadhead, 1959; Yang, 1971; Schaefer, 1989, 1991), length-frequency and catch-and-effort analysis (Yabuta et al., 1960; Kamimura and Honma, 1963), tagging experiments (see below), spawning studies (Section 5.2), and other studies dealing with various aspects of the life history of yellowfin. A considerable amount of genetic research has been carried out in an attempt to develop a direct method to discriminate yellowfin subpopulations (Suzuki, 1962; Barrett and Tsuyuki, 1967; Sprague, 1967; Fujino and Kang, 1968a and 1968b; Fujino, 1970; IATTC, 1971–74 and 1976–80). Some of these studies have produced inferences contradictory to those obtained from the indirect sources; however, due to sampling difficulties, limited geographical coverage and the need for further systematic follow-up investigations, they are still considered to be preliminary in nature (Suzuki et al., 1978; IATTC, 1980). Therefore, determination of the existence or non-existence of subpopulations of Pacific yellowfin has yet to be made.

On the basis of the available indirect data, however, Suzuki et al. (1978) drew the following conclusions with respect to population structure:

1. Although no definite geographical break in the continuity of yellowfin distribution across the Pacific is indicated, the extent of movement appears to be insufficient for there to be much intermingling among yellowfin of the eastern and central Pacific, and probably of the western and central Pacific as well.

2. Three stocks, which are more or less independent, are envisioned: a western Pacific stock which inhabits the area between about 120°E and 170°W; an eastern Pacific stock inhabiting an area corresponding to the IATTC's yellowfin regulatory area (CYRA) (Figure 4); and a central Pacific stock inhabiting the area between the western and eastern stocks. (A stock was defined to mean “… an exploitable subset of the population existing in a particular area and having some uniqueness relative to exploitation.”)

3. The three stocks may well be composed of subpopulations which will have to be discriminated by more direct methods. Should the existence of subpopulations be established, however, it is possible that even direct genetic studies will not demarcate their spatio-temporal boundaries.

In his discussion of the contribution of life history data to the analysis of population structure, Bayliff (1983) noted that the average size of yellowfin within the CYRA was reduced by fishing, whereas this effect was hardly apparent in the area between the CYRA and 150°W. This observation suggested that yellowfin in the two areas may be relatively more independent than those in the central and western Pacific. However, more recent results (IATTC, 1991a) indicate that the apparent difference in averages may have been caused by the scarcity of data in each area due to the effects of regulated and unregulated periods of fishing. In the absence of regulations since 1980, the trends in the average sizes have become more similar as fishing effort in the two areas has become more evenly distributed in space and time. This subject is explored more fully in Section 12.

In a recent study involving discriminant analyses of morphometric characters, Schaefer (1989) concluded that eastern Pacific yellowfin sampled from north of 15°N–20°N were different from those sampled from south of 15°N–20°N. Somewhat distinct regional groups were suggested by the absence of any clinal relationship between the characters and latitude or longitude. On a larger oceanic scale, yellowfin samples from Mexico, Ecuador, Australia, Japan and Hawaii showed significant differences in gill-raker counts and morphometrics, suggesting that these different regions represent separate groups (Schaefer, 1991).

Figure 4. The Inter-American Tropical Tuna Commission's Yellowfin Regulatory Area (CYRA). (Reproduced from IATTC, 1991a).

6.2 Distribution

Yellowfin are distributed worldwide, occurring in the tropical and subtropical waters of the Indian, Pacific, and Atlantic Oceans, and in all the warm seas of the world except the Mediterranean Sea. The coastal limits of their distribution in the Pacific were documented by Rosa (1950). He reported that in the eastern Pacific, yellowfin range from Point Conception, California, to San Antonio or Talcahuano, Chile, and in the western Pacific they are distributed from Hokkaido, Japan, through the Indonesian Archipelago, to Cape Howe, Australia, and New Zealand. Blackburn (1965) suggested that the latitudinal limits of yellowfin are about 35°N and 33°S in the eastern Pacific, and 40°N and 35°S in the western Pacific. Kamimura and Honma (1963) stated that yellowfin are distributed in practically every part of the Japanese longline fishing grounds in the Pacific, which at the time extended from about 40°N to about 40°S, and from the Asian coast eastward along the equator to about 100°W. The longline fishery has since expanded further into the eastern Pacific (Figure 5). Catch records from this fishery and the purse-seine fishery indicate the relatively continuous distribution of yellowfin within its range (Figure 6), which extends across the Pacific and is roughly enclosed by the 40°N and 40°S latitudes (Calkins and Chatwin, 1967, 1971; Shingu et al., 1974; Calkins, 1975; Fisheries Agency of Japan, 1974–77; IATTC, 1980, 1989; Suzuki et al., 1978; Miyabe and Bayliff, 1987).

Figure 5. Geographical expansion of the Japanese longline fishery (solid curves) and the surface fishery in the eastern Pacific Ocean (dotted curves). Numerals denote calendar years. (Reproduced from Suzuki et al., 1978).

6.3 Movement

In the eastern Pacific, yellowfin movement was initially studied by Blunt and Messersmith (1960) and Schaefer et al. (1961). A later report by Fink and Bayliff (1970) included the data on all tagged fish prior to 1965. They concluded that from 1952–1964 the fishery operated on two main groups of yellowfin: a northern group off the west coast of Baja California, in the Gulf of California, and around the Revillagigedo Islands; and a southern group from the Tres Marias Islands to northern Chile. The seasonal movements of the two groups appeared to be largely in a coastal direction with some mixing taking place between them. The lack of tag recoveries by the Japanese longline fleet operating outside the CYRA and on grounds partially overlapping the nearshore surface fishery, suggests that there was no large-scale offshore-inshore movement of yellowfin during the 1952–1964 period (Joseph et al., 1964). Schaefer et al. (1961) also pointed out that fish tagged near fishing banks tended to disperse very slowly.

Figure 6. An example of a monthly distribution of average catch rates for yellowfin caught by Japanese longline boats, for 1967–72 combined. (Reproduced from Suzuki et al., 1978).

As the surface fishery began moving progressively offshore during the late 1960s, more fish were recaptured offshore. Nevertheless, no tagged yellowfin released in the eastern Pacific have been recovered outside the eastern Pacific (Bayliff, 1984), and the concept of three, semi-independent stocks (Section 6.1) in the Pacific Ocean has not been refuted. Studies conducted off southern Mexico, northern Central America and offshore (Bayliff and Rothschild, 1974; Bayliff, 1979, 1984) indicate that differences in the catches of smaller fish in the inshore and offshore areas were due mainly to differences in vulnerability. In addition there were no pronounced tendencies for movements in an east-west or north-south direction even though the movements were not random. Although the overall implication is that yellowfin do not move in excess of several hundred miles (see Table 2, Miyabe and Bayliff, 1987), exceptions exist and a few interesting examples appear in Table 3.

7. NATURAL MORTALITY

Hennemuth (1961b) analysed length-frequency data to obtain an estimate of 1.72 for the total annual instantaneous mortality rate (Z) of yellowfin in the eastern Pacific for 1954–59. Using the coefficient of catchability estimated by Schaefer (1957) and an estimate of fishing effort, he estimated the annual instantaneous fishing mortality rate (F) to be 0.95 and the annual instantaneous natural mortality rate (M) to be 0.77, with 95-percent confidence limits of 0.64 to 0.90. Hennemuth regarded 0.80 as the most probable value of M and 0.60 to 1.00 as the “extreme possible” values. Schaefer (1967), however, considered the lowest and highest probable values to be 0.55 and 1.05, respectively. Murphy and Sakagawa (1977) evaluated the published estimates of M and concluded that the values around 0.80 were the “best” estimates. Bayliff's (1971) result of M <2.0 does not dispute this conclusion although the estimate, derived from tagging, was apparently affected by the temporal variation in vulnerability and the failure of tagged and untagged fish to mix completely. Francis (1977), however, based on a simulation model of the dynamics of the eastern Pacific surface fishery, suggested that the value of 0.80 was too high. He used this rate and Hennemuth's (1961b) lowest value of M = 0.60 as input parameters to modified versions of an earlier model (Francis, 1974) and found that 0.60 was a better estimate in validating the versions which best mimicked the fishery. Nevertheless, until recently, the value of M = 0.80 was used by the IATTC (IATTC, 1989) in age-structured assessment models because the interpretation of data from the fishery was most compatible with this value. As a result of the virtual disappearance from the fishery of females larger than 140 cm, different values of M have been applied to the sexes beginning with stock assessments in 1989. This subject is discussed more fully in Section 12.

TABLE 3. Unusual returns of yellowfin tagged in the Pacific Ocean (IATTC, 1980, 1982, 1983).

8. OCEANOGRAPHIC FEATURES ASSOCIATED WITH THE SPECIES

The principal circulation pattern in the upper waters of the Pacific Ocean (Figure 7) is driven by the easterly trade winds and is not symmetric with respect to the equator (Pickard, 1968). The well-developed surface components of the system include a westward-flowing South Equatorial Current (SEC) between 8° to 10°S and 3°N, a westward-flowing North Equatorial Current from about 8° to 20°N, and between them a narrow North Equatorial Counter Current flowing to the east. An additional easterly component that normally lies beneath the SEC (IATTC, 1984) consists of the generally weaker South Equatorial Counter Current (SECC). This system can be traced from the Gulf of Panama almost to the Philippines, a distance of some 12,000 km. The relationship of this circulatory pattern to other, eastwardly-flowing, subsurface currents is demonstrated in Figure 8 by means of a transect from Hawaii to Tahiti (Wyrtki and Kilonsky, 1984).

Figure 7. Pacific Ocean surface circulation. (Reproduced from Pickard, 1968).

Figure 8. Areas occupied by the main zonal currents between Hawaii and Tahiti in the upper 400 m. Dark shading is westward flow, light shading eastward, and blank areas have zonal speeds less than 2 cm/s. Abbreviations: NEC - North Equatorial Current; SEC - South Equatorial Current; NECC - North Equatorial Counter Current; SECC - South Equatorial Counter Current; NSCC - North Sub Counter Current; SSCC - South Sub Counter Current; UC - Under Current; EIC - Equatorial Intermediate Current. (Reproduced from Wyrtki and Kilonsky, 1984).

Of the three major subsurface currents, the core of the Equatorial Undercurrent, or the Cromwell Current, is located in the thermocline during normal, or non-El Niño periods, and extends approximately 10,000 km westward from the Galapagos Islands. During these periods the decrease in the sea-surface temperature from west to east is associated with a shoaling of isotherms (Figure 9), and a gradual rise in the depth of the thermocline from about 150 m in the western Pacific to 40 m or less in the eastern Pacific (Philander, 1990). Figure 9 also illustrates that the greatest change in the thermocline depth occurs in the central and eastern Pacific regions.

Figure 9. Temperature (°C) as a function of depth and longitude along the equator in the Pacific Ocean, as measured in 1963 by Colin et al., (1971). (Reproduced from Philander, 1990).

During El Niño episodes all of the tropical currents in Figure 7, including the California and Peru (Humboldt) currents in the eastern boundary, undergo changes in their normal positions and strengths (IATTC, 1984). Following a weakening of the southeasterly trade winds in the eastern part of the south Pacific, some of the principal changes that take place in the eastern Pacific include a rise in the sea surface due to a deceleration of the hydrospheric circulation in the south Pacific Ocean, elevated coastal and equatorial temperatures together with a movement of warmer surface water over the normally upwelling regions off Ecuador and Peru, and a deepening of the thermocline (Miller and Laurs, 1975; Joseph and Miller, 1988). In addition, while the SEC weakens, the volume of the SECC may increase and the current is then frequently found at the surface between 5° –10°S. The Cromwell Current also weakens, and during a particularly strong El Niño, such as in the last quarter of 1982 and into 1983, the current is difficult to locate in transect data at the equator. Collectively, these events are thought to have an effect on the recruitment of yellowfin and the vulnerability of yellowfin and skipjack to capture.

As Figure 10 illustrates, strong, positive recruitment anomalies in 1971, 1974, 1978 and 1985 were preceded by pronounced El Niño conditions in 1969, 1972, 1976, and 1983 (Joseph and Miller, 1988). Weaker El Niño events also occured during the third quarter in 1982 and throughout 1987. Consequently from 1982 through 1986, waters in the eastern Pacific were warmer than usual and recruitment during the 1984–1988 period was the greatest on record for a five-year interval (IATTC, 1989). Although the causal relationship between enhanced recruitment and El Niño conditions…if it exists…is not understood, the survival of young fish and the rates at which eggs, larvae or postlarvae are retained in the eastern Pacific are probably increased due to the diminished flow of the altered current patterns.

Figure 10. Anomalies from average recruitment of yellowfin, in millions of fish, in the eastern Pacific Ocean. (Reproduced from Joseph and Miller, 1988).

El Niño episodes also affect the vulnerability of yellowfin and skipjack tunas to purse-seining operations. This has been studied in three heavily-fished areas: the inshore coastal area off Ecuador and northern Peru; the Costa Rica Dome (CRD), about 200 miles west of Nicaragua and Costa Rica; and the triangle between Cape San Lucas, Cape Corrientes (Mexico) and the Revillagigedo Islands (IATTC, 1989). In particular, for yellowfin within a 300-mile radius of the CRD, the trends in catch per day's fishing (Section 11), or CPUE, the successful-set ratio (SSR), i.e. the ratio of successful sets to all sets, and the availability-vulnerability (AV) index are shown in Figure 11. The index is calculated as CPUE/(1 - SSR) so that an increase in either CPUE or SSR will also increase the index. The 1983 data reflect conditions during the most extensive and severe El Niño in this century (IATTC, 1988). Reduced upwelling, that persisted from November 1982 to May 1983 in the CRD region, reduced the vulnerability of the fish to capture, and depressed the CPUE and hence the AV index. During the last half of the year the El Niño began to weaken, but not sufficiently to allow the fishery to recover. At the same time a major portion of the fleet departed for the western Pacific. During 1984 the average CPUE recovered to 13.5 mt/day compared to 2.5 in the previous year. In contrast to 1983, the mild El Niño episode of 1987 affected only the region south of 5°N. The thermocline was 15 m above normal at the CRD, but 20 m deeper south of 5°N and east of 135°W. This situation apparently increased the CPUE to above-normal levels throughout most of the year and, combined with an elevated SSR, these conditions produced an AV index at least twice the long-term 1975–88 average. The oceanic conditions of 1987 persisted through 1988, with additional upwelling and a shallower thermocline. In response to these favourable conditions, the AV index maintained a high level through increases in both the CPUE and SSR.

Figure 11. Monthly mean values of CPUE, effort, successful set ratio, and availability-vulnerability index for yellowfin at the Costa Rica Dome. (Reproduced from IATTC, 1989).

There is little information describing the effect of current patterns and other hydrodynamic activity on the fishing success of yellowfin other than in association with the thermocline. For example, Suda and Schaefer (1965a) confirmed earlier observations that in the equatorial zone there is a gradual increase in the average weight of longline-caught yellowfin in the direction from the western to the eastern Pacific. They speculated that the size gradient was a product of gear selectivity, size stratification within the depth habitat of yellowfin, and the easterly shoaling of the thermocline (see above). They recognized that while yellowfin of all sizes occurred throughout the mixed layer and the upper part of the thermocline, larger fish may be more abundant in the lower mixed layer and the upper thermocline. In addition, since conventional longlines fished at a maximum depth of about 160 m during the early 1960s, the gear became increasingly effective in catching larger fish in an easterly direction due to the decreasing depth of the thermocline. A subsequent comparison by Suzuki et al. (1978), of the sizes of fish caught by purse-seine and longline gears operating in overlapping areas in the eastern Pacific, produced no evidence of vertical size stratification. However, they concluded that the easterly cline in the central and eastern Pacific in the size of longline-caught yellowfin was due to gear selectivity in conjunction with the shoaling thermocline.

In retrospect, it is possible that the vertical size stratification observed by Suda and Schaefer (1965a) was real and a product of the operational depths of the two gear types at that time. For instance during the mass conversion of baitboats to purse seiners from 1959 to 1961, if the working depth of a typical, 73-m net (McNeely, 1961) was approximately 40 m, or 54 percent of its stretched depth, and that of a conventional longline was 50 to 100 m (Suzuki et al., 1977), then the two gears may have selected somewhat different sizes without interacting. During the 1970s, however, the working depth of purse-seine nets increased to about 70–80 m, and presently they are up to 110 m (19 strips) and 160 m (28 strips) in the eastern and western Pacific, respectively (J. Lewis, Casamar Inc.; pers. commun). In this situation it is unlikely that the comparative catches of these gear types can help resolve the issue of whether yellowfin are stratified by size in the water column.

During the decade of the 1970s, scientists of the IATTC and the U.S. National Marine Fisheries Service (NMFS), Southwest Fisheries Center, collected and analysed many hundreds of expendable bathythermograph and surface observations taken aboard tuna boats in the eastern Pacific. From these data the hypothesis evolved that yellowfin in schools without dolphins were more vulnerable to purse seines when the 23°C and 15°C isotherms, were close to the surface and the gradient between these isotherms, or strength of the thermocline, was most pronounced. The hypothesis was based on the assumption that yellowfin do not attempt to escape from under the net when the bottom of the net is located in or below the bottom of a strong thermocline. Discriminant analysis revealed that, of several environmental parameters tested, either a shallow 23°C isotherm depth or a strong gradient contributed significantly to increased catch rates of “schoolfish”, or fish associated only with other fish. The depths of either the 23°C or 15°C isotherms, however, had little effect on the catch rates of yellowfin associated with dolphins. In this case a strong gradient led to the greatest catch rates. Collectively, these results suggest that the depth of the yellowfin habitat may be determined to a large extent by the vertical temperature structure, or related properties, such as oxygen content (IATTC, 1982). Acceptance of these results must also be tempered by the fact that changes in market needs and in fishing strategies can mask the environmental effects on fishing effort. There have also been numerous occasions when the presence of strong gradients and a shallow 23°C isotherm have not resulted in improved catches. The apparent abundance of schoolfish may therefore be more closely related to other factors, such as ocean currents which aggregate or disperse food (IATTC, 1981; Sund et al., 1981).

The use of sonic tagging devices has augmented, but not changed, the earlier belief that yellowfin occupy the mixed layer and the upper portion of the thermocline. Carey and Olson (1982) attached acoustic transmitters to three, 87–98 cm yellowfin in the Clipperton Island-Gulf of Panama region and found that their primary orientation was to the thermocline, with frequent vertical excursions above and below this reference. Solitary fish also tended to stay near the thermocline, while those which joined schools moved up into the mixed layer. During daytime the fish were found at greater depths than at night, but collectively they spent little time at the surface. Similar experiments by Holland et al. (1990) near the Hawaiian Islands indicate that the mixed layer and the first degree (°C) of reduced temperature at the top of the thermocline accounted for 68 percent of the daytime distribution of four, 54–74 cm yellowfin, and 77 percent at night. The introduction of deep longlines, whose maximum hook depth can penetrate to 160 m or 250 m, depending on the area fished, increased the hook rate of bigeye tuna, but not that of yellowfin (Hanamoto, 1974; Suzuki et al., 1977). This finding supports the view that the principal habitat of yellowfin is in the upper boundary of the thermocline and mixed layer. Deep forays to 168 m and 464 m (Carey and Olson, 1982), where the oxygen partial pressures were estimated to be 40 to 60 percent and 20 percent of air saturation, respectively, suggest that yellowfin are tolerant of low oxygen levels. However, Green (1967) noted that, based on typical data from the EASTROPIC cruises, the oxygen content in the mixed layer was 4 to 6 ml/l followed by a sharp decline to 1 ml/l just below the thermocline. Bushnell et. al. (1990) found that yellowfin responded significantly in terms of gape, ventilation volume and heart rate when hypoxic oxygen levels reached 2.7 to 3.3 ml/l. Consequently, it is not clear whether the orientation of yellowfin to the top of the thermocline is a preferential response to temperature, or if it represents the maximum depth at which a minimal amount of oxygen is consistently available.

9. INTERACTION WITH OTHER SPECIES

9.1 Dolphins

In the eastern Pacific a situation exists in which predominantly moderate-to-large sizes of yellowfin frequently associate with one or more species of dolphins. Fishermen have long recognized this relationship and purposely search for the surface activity of dolphins because it may indicate the presence of sub-surface yellowfin. Seabirds, such as the sooty terns (Sterna fuscuta), wedge-tailed shearwaters (Puffinus pacificus) and boobies (Sula spp.), are frequently found over dolphin schools, and extend the fishermen's horizon of detection (Au et al., 1979). Exploitation of the yellowfin-dolphin association began in earnest by purse seiners following the conversion of most of the baitboat fleet to purse seiners during 1959–1961. At that time about 62 percent of the catch was taken in association with dolphins (Perrin, 1969). Concern for dolphin mortality incidental to the fishing process has subsequently led to legislation in the USA to monitor their abundance and mortality, and to reduce mortality through application of the Marine Mammal Protection Act (MMPA) of 1972 (U.S. Dept. of Commerce, 1977). The NMFS implemented these provisions, including the establishment of annual quotas for the different dolphin species, enforcement of the regulations, and placement of observers aboard USA vessels. The research programme conducted by the NMFS also focussed on the essential factors of population estimates, reproductive rates, and gear and procedural modifications to reduce mortality (U.S. Dept. of Commerce, 1978). The MMPA was strengthened in 1984 by recommendations that: 1) the maximum dolphin mortality rate inflicted by vessels of a foreign nation conform to that of USA vessels, under penalty of embargo of commercial fish or fish products, and 2) that these restrictions also apply to intermediary nations. In 1977 the IATTC initiated a programme that involved the co-operation of the international purse-seine fleets operating in the eastern Pacific. The plan is similar to that conducted by the NMFS, but lacks enforcement provisions. Attributes of the programme that are oriented toward conservation include gear research, the use of educational workshops to disseminate information on safe practices, and behavioural research surrounding the interaction of dolphins and yellowfin tuna (IATTC, 1978). In 1990 all of the principal tuna canners in the USA announced that they would no longer purchase tunas caught in association with dolphins. The consequences of this action are presently unknown.

Yellowfin are found together primarily with spotted dolphin (Stenella attenuata), to a lesser extent with spinner (S. longirostris) and common (Delphinus delphis) dolphins, and relatively seldom with striped dolphin (S. coeruleoalba) and other species (Hammond, 1981). These degrees of involvement may indicate that the basis of the association is partly due to a shared or preferred food source. For example, Perrin et. al. (1973) found that the food items in the stomachs of yellowfin and spotted dolphin caught in the same nets overlapped considerably, with ommastrephid squids representing the most common forage item. The diets of the spinner dolphin and yellowfin, on the other hand, appeared to be quite different. With common dolphins the association with yellowfin seems to have seasonal, areal and yearly components. If a common food preference is one ingredient in the linkage, the disadvantages of the implied competition must be offset by one or more beneficial elements. Hammond (1981) explored the interactive possibilities between two populations, including competition, parasitism and commensalism, as well as the implications of tunas seeking dolphins or the reverse. While there is an overriding need for observational or experimental evidence to evaluate these possibilities, there is nevertheless some indication that, in addition to the benefits of schooling, the association may also be based on enhanced protection from mutual predators such as sharks and small whales.

9.2 Skipjack and Bigeye Tunas

In studying the size composition of tuna schools in the eastern Pacific, Orange et. al. (1957) and Broadhead and Orange (1960) noted that yellowfin and skipjack occurred in mixed schools. In general, about 10 percent of the purse-seine catch and 35 percent of the baitboat catch resulted from mixed schools, with the remainder in each case being made up of fish from pure schools of either species. The relative occurrence of mixed schools appeared to be greatest when the proportion of pure yellowfin and skipjack schools were evenly divided. On average, the yellowfin and skipjack in mixed schools were similar in size, but the sizes of yellowfin in pure schools appeared to be larger and more variable. The skipjack from pure and mixed schools tended to be more alike.

Yuen (1963) challenged the implied species interaction in mixed schools with evidence from underwater observations and photography of baitboat fishing. The yellowfin and skipjack in seemingly composite aggregations actually maintained a significant degree of separation into schools or species groupings, and the random response of individuals to the proffered food created the illusion of complete mixing. The conclusion was drawn that while yellowfin and skipjack prefer to school by species, their co-occurrence is probably the congregation of separate schools responding to the common attraction of an external stimulus, such as food.

The amount of bigeye tuna caught annually by longline gear in the eastern Pacific has varied by a factor of two, from about 36,300 to 72,600 mt, in the interval from 1961 to 1980 (IATTC, 1991a). Over a slightly longer period, from 1961 to 1988, the modal catch of this species by the surface fleet was less than 907 mt, with a range from 68 to 15,400 mt. Differences of this magnitude between gear types suggest that the habitat preference of the two species preclude any large degree of interaction.

A recent study (IATTC, 1991b) indicated that the possibility of misidentifying bigeye tuna as yellowfin does not occur to an appreciable degree in the eastern Pacific surface fishery. During the interval from April, 1987, to September, 1989, observers aboard USA purse seiners recorded that, in floating-object and school sets, 46 mt of bigeye were caught and loaded compared to 17,155 mt of other tunas, or a tonnage ratio of 0.0027. The observers had received special training to identify bigeye (> 30–35 cm) in the commercial catch. Based on logbook entries, cannery unloading weights and length-frequency samples, the comparable ratio for all other USA purse seiners in the same set-type, time-area strata was 0.0026. Since the two ratios are essentially equal, it does not appear that the bigeye catch by USA vessels is misrepresented by the current methods of evaluation. Non-USA and USA purse seiners fish in the same overall area in the eastern Pacific, but the non-USA vessels tend to capture greater amounts of bigeye near the Galapagos Islands and south of the equator. Consequently, the tonnage ratio of bigeye to all other tunas for the non-USA vessels was 0.0066. However, since the same sources of information are used to estimate the landings for both segments of the fishery, there is no reason to suspect that bigeye landings are misrepresented for the purse-seine fleet as a whole.

10. GENERAL DESCRIPTION OF THE FISHERY

Prior to the late 1950s the surface fishery for yellowfin in the eastern Pacific was dominated by baitboats operating in coastal waters and in the vicinity of offshore islands. The conversion of most of the baitboats to purse seiners in the late 1950s and early 1960s fostered and accelerated the seaward expansion of the fishery. In 1968 the purse-seine fleet began fishing in the area outside the CYRA (Figure 4), and by 1974 fishing was conducted as far west as 150°W (Calkins, 1975; Peterson and Bayliff, 1985). As a result of these changes the number of baitboats declined from 204 to 36 during the 1950–88 period, while their capacity decreased from about 35,380 to 2,810 mt. During the same interval the number of purse seiners increased from about 67 to 182, with an increase in capacity from approximately 7,160 to 134,300 mt (IATTC, 1989, Table 4). Fishing pressure on yellowfin also increased due to the eastward expansion of the Japanese longline fishery that began in 1948. Catches to the east of 130°W were first reported in 1956, and trans-Pacific longlining was essentially completed by 1962 (Figure 5) (Suda and Schaefer, 1965a). With the exception of the region surrounding French Polynesia, the fishery in the eastern Pacific Ocean (EPO) now comprises the area to the east of 150°W between the 39°N and 35°S latitudes. In addition to the gear types just mentioned, lesser amounts of yellowfin are also caught by coastal dayboat operations involving smaller vessels (bolicheras and jigboats). By 1990 vessels of at least 10 nations had participated in the yellowfin surface fishery and three in the longline fishery.

During the 1979–87 period the average catch in the CYRA by the surface fleet and the Japanese longline fleet was 162.6 × 10³ mt (range: 82.8–247.6), and the preliminary estimate for 1988 was 267.6 × 10³ mt. For only the surface fleet in the region between the boundary of the CYRA and 150°W, the average catches for the 1979–87 period and in 1988 were 22.2 × 10³ mt (range: 12.2–38.7) and 20.7× 10³ mt, respectively. Total catches in the EPO in 1988 were therefore at least 288.3 × 10³ mt, the greatest amount on record, and 15.8 × 10³ mt in excess of the previous record in 1987 (IATTC, 1989). The distribution of catches for the 1979–87 and 1988 surface segments of the fishery appear in Figures 12a and b, respectively.

“The need for regulation of the fishery for yellowfin was first apparent in 1961, but the concerned governments participating in the IATTC were not able to implement regulations until 1966. Thereafter regulations were in force through 1979. For 1980 and subsequent years [except for 1987] the IATTC has recommended quotas, but the countries which participate in the fishery have not been able to agree on their implementation” (Peterson and Bayliff, 1985).

As an oversimplification, in years in which regulations were in effect, the fleet operated throughout the EPO during the initial part of the year until established quotas were met, and then outside the CYRA for the balance of the year. Under these conditions it is difficult to disentangle what may be the seasonal component of fish movement from the obligatory movement of the fleet. Even in the absence of quotas after 1979, it is also possible that the observed distribution of CPUE is the result of the confounding effect of fish and fleet movements. Nevertheless, Punsly (1987) noted certain trends in the distribution of sets on schoolfish, and fish associated with dolphins or logs in several season-area (SA) strata (Figure 13). For example, during the 1970–1985 period, which included both regulated and non-regulated years, the schoolfish catch rate was greatest in February through April near the Gulf of Panama (SA 8), second in rank during November through March near the tip of Baja California (SA 20), and third during September and October off Ecuador (SA 6). The catch rates on fish associated with dolphins on the other hand, were greatest during November and December in the area offshore south of the equator (SA 3), and second off the tip of Baja California (SA 20): Catch rates on fish associated with floating objects were greatest during September and October off Central America and southern Mexico (SA 13), and second near the Gulf of Panama (SA 8).

11. TRENDS IN FISHING EFFORT, CATCH AND CATCH PER UNIT OF FISHING EFFORT

The catch per day's fishing (CPDF), as an index of yellowfin abundance, has been revised several times during the history of the fishery. When baitboats dominated the surface fleet in the 1950s, the catch and CPDF data for vessels of different size classes were standardized (S) to calculate the CPSDF for Class-4 (182–272 mt capacity) baitboats. Following the conversion of the fleet to purse seining by the early 1960s, Class-3 (93–181 mt) seiners became the new standard that persisted until 1967. As early as 1962, however, the CPSDF was adjusted by factors related to the SSR to take increases in fleet efficiency into account. By 1974, as a result of increased vessel construction that began in the mid 1960s (Table 4), the share of the fleet capacity occupied by Class-3 seiners declined to 3 percent while that of Class-6 (≥ 364 mt) seiners rapidly increased to 81 percent. For a period of time, indices based on standards for both classes of purse-seine vessels were used to monitor relative abundance, although increased reliance was placed on the CPDF derived for the dominant, Class-6 seiners. By 1983 this class represented almost 90 percent of the surface fleet capacity, and the raw effort expended by this class alone has since been used to measure the CPDF. Additional measures of relative abundance were developed by Pella and Psaropulos (1975), Allen and Punsly (1984) and Punsly (1987), utilizing tons caught by purse seiners per hour of searching. The rates were standardized by a weighted, generalized linear model that took into account the significant effects of vessel speed, season-area and whether the yellowfin were caught with dolphins, floating objects or skipjack. Biomass indices (Section 12.2) have also been constructed by the application of cohort analysis to catch and length-frequency data. The trends in these three measures of abundance appear in Figure 14 (IATTC, 1991a). Compared to the indices based on searching time, the CPDF for Class-6 seiners underestimated the relative apparent abundance during the late 1970s and overestimated it during the early 1980s. The CPDF and searching-time indices show a declining population after 1986, with a partial recovery in 1989, while the biomass index reflects a stable period after 1985, followed by a decline in 1989.

Figure 12. Average annual catches of yellowfin in the EPO during 1979–87 (a), and catches in the EPO in 1988 (b), for all purse-seine trips for which usable logbook data were obtained. (Reproduced from IATTC, 1989).

Figure 13. Time-area strata used to standardize catch rates to develop abundance estimates based on searching time. (Reproduced from Punsly, 1987).

In Table 4 the historic trends in yellowfin catch, effort, CPDF, fleet capacities and recruitment contain several periods of special interest. Following the introduction of fishing regulations in 1966 and the first ventures by purse seiners outside the CYRA in 1968, the CPDF recovered to its highest level (15.5 mt/day) in 1969 since the earliest days of the fishery. This was due in part to the fishery's concentration on larger fish, and 1967 to grow and contribute to that part of the population (IATTC, 1981). However, the interaction of three factors: below-average recruitment from 1969 to 1972 and in 1976 and 1977, accelerated construction of vessels in the early 1970s, and exploitation of both large and small fish, ultimately reduced abundance to its lowest levels by 1981–82. The extremely large recruitment of 1973 temporarily raised the abundance of small fish and thereafter contributed to the biomass of larger fish in 1975 through 1977. In fact, the total catch in 1976, 251.4 × 10³ mt, was the largest on record at that time, but the pulse of this strong recruitment moving through the fishery was not sufficient to halt the downward trend in abundance. This condition, coupled with the onset of the severe El Niño of 1982–83, led to reduced catches and an exodus of fishing vessels to the western Pacific. During the ensuing period from 1984–86, several factors were thought to contribute to the recovery of the fishery, including: (1) above-average recruitment following the El Niño event; (2) declining effort from 1982–86, although some vessels began returning from the western Pacific in 1984; (3) reduced effort on small yellowfin due to a reduction in prices for skipjack and small yellowfin; and (4) an increase in the effort directed toward larger-than-average fish (IATTC, 1989). The elevated sea-surface temperatures resulting from the 1982–83 El Niño were extended by the weaker El Niño of 1987 and may have contributed to the exceptional recruitment of 1987. In addition, as in 1985–87, it appears that the fishery began to concentrate on large fish again in 1989, following a brief period in which small fish were also captured in 1988. These favourable conditions no doubt contributed to the succession of increased catches that culminated in records of 301.7 × 10³ mt in 1988 and an estimated 296.0 × 10³ mt in 1989, but such large catches may have also initiated a decline in abundance (Figure 14).

TABLE 4. Catches of yellowfin, effort and catch per days fishing (CPDF) by the surface fleet and longline fleets in the eastern Pacific Ocean. The historic carrying capacities and numbers of vessels (No.) in the surface fleet, and the initial number of fish in the X and Y cohorts recruited at 30 cm in length are also included. Weight units are short tons (st). (Adapted from IATTC 1991a).

* effort for all surface vessels in class-6 (>400 st) purse-seine units.
** for class-6 purse seiners.
# preliminary data for 1988.
## all data for 1989 are preliminary.
@ Quotas established but not implemented from 1980 through to present.

Figure 14. Three indices of abundance of yellowfin in the eastern Pacific. (Reproduced from IATTC, 1991a).