Tagging Studies on the Shortfin Mako Shark (Isurus oxyrinchus) in the Western North Atlantic

Atlantic coast of North America and in the Gulf of Mexico in the 28-year period between 1962 and 1989. The sharks were tagged by volunteer fishers, scientists aboard research vessels, and US Foreign Fisheries Observers on foreign vessels fishing in US waters. Altogether, 231 recaptures (9-4% of releases) of these tagged sharks have been reported by fishers from 16 countries. The maximum time at liberty is 8-2 years, and the maximum straight-line distance between tag and recapture localities is 2452 nautical miles. One tag was recovered from the European side of the Mid-Atlantic Ridge. Mako shark distribution and migratory patterns in relation to water temperature are discussed for the western North Atlantic.

The shortfin mako shark, Isurus oxyrinchus Rafinesque (1809), is a member of the family Lamnidae, which is comprised of three genera and five species. Collectivel~ known as the mackerel sharks, the species reported from the western North Atlantic Ocean are the white shark (Carcharodon carcharias), the porbeagle (Lamna nasus), the longfin mako (1surus paucus), and the shortfin mako (1surus oxyrinchus) (Compagno 1984). In this report, the common name 'mako shark' refers to L oxyrinchus because no information on L paucus is included.

The mako shark is distributed throughout the temperate and tropical zones of the world's oceans. Females reach a maximum size of 394 cm total length (TL) and mature at about 280 cm TL. Males mature at 195 cm TL and reach a maximum length of 284 cm TL (Stevens 1983; Compagno 1984). Makos are ovoviviparous, producing 4-16 pups per litter. Pups are approximately 70 cm TL at birth (Stevens 1983) after a gestation period of 812 months (Pratt and Casey 1983). Makos feed on a wide variety of fishes and cephalopods in different parts of the world and are one of the few predators of large swordfish (Xiphias gladius) (Stillwell and Kohler 1982; Compagno 1984).

Mako sharks are caught in commercial fisheries in the Atlantic Ocean, primarily as a by-catch in longline fisheries directed at tunas and swordfish. The total catch of makos is unknown because landings are often not reported and the sharks, as a nontarget species in the catch, are discarded or released at sea. The Food and Agriculture Organization of the United Nations began listing mako shark landings in 1987. Total annual landings reported for the Atlantic between 1987 and 1989 were only 6-7 tonnes (t), with only the United States

reporting (Anon. 1991). Nevertheless, makos are taken as a by-catch in the longline fisheries of several nations, including Cuba, Denmark, Venezuela, Korea, Taiwan, Japan, Spain, Canada and the United States (the last four being the principal countries involved in these fisheries). Hoey (1983) showed that mako sharks were closely associated with swordfish over the geographical range covered by the US swordfish longline fishery. He analysed catch data from tuna and swordfish fisheries in the western North Atlantic and found that mako sharks represented a much higher percentage of the by-catch in the swordfish fishery (2.9%) than they did in the US (I - 6%) and Japanese (I - 3%) tuna fisheries.

The US and Canadian longline fisheries for swordfish in the Atlantic began in the early 1960s. For the first few years, fishing effort was confined to the area between Cape Hatteras, North Carolina and Georges Bank (35-40'N) and in the vicinity of the Grand Banks. Initially, US and Canadian swordfish longline gear was composed of multistrand line that evolved from Japanese tuna gear (Hoey and Casey 1986). After 1978, however, many US vessels entering the swordfish fishery from the southern states changed to monofilament line and modified the gear in other ways by setting fewer hooks at greater depths and employing artificial-light sticks. Since the 1970s, the US swordfish fishery has expanded and currently extends over much of the western North Atlantic from the Grand Banks to the equatorial zones off South America, including the Gulf of Mexico and the Caribbean Sea (Hoey and Mejuto 1991). Although US longline fishers occasionally attempt to target mako sharks, there is currently no US commercial fishery directed at makos in the Atlantic.

The Japanese longline fishery in the Atlantic began in 1956 and covered almost the entire Atlantic by the late 1960s. The initial target species was yellowfin tuna, then albacore, and by the 1980s the fishery was directed at bigeye tuna. Yao (1988) plots catch rates of swordfish and the Japanese longline effort for the entire Atlantic. He notes that the deeper longlines used for bigeye tuna increased the catch per unit effort (CPUE) for swordfish in tropical waters. The distribution of Japanese fishing effort and swordfish CPUE from 1969 to 1972 shows some fishing effort in every 5' grid square in the North Atlantic south of 40'N. From 1983 to 1985, the Japanese effort in the western North Atlantic was confined to the Caribbean and to a northern region between 35 and 45'N. In the same period, CPUE data for swordfish were reported for most of the eastern Atlantic between 20'S and 40'N (Yao 1988).

The Spanish longline fishery in the eastern Atlantic is directed primarily at swordfish. The Spanish fleet employs a standard type of surface longline, with varying numbers of hooks used per set (Mejuto and Garc6s 1984). The gear is similar to the early US swordfish longline. Prior to the mid-1980s, the distribution of the North Atlantic Spanish swordfish longline fishery was confined to the area between theAzores and the Iberian peninsula (35-45-N,8-30-W). For 1983 and 1984, the first years for which by-catch data are available, the numbers of mako sharks taken in this fishery were 5525 (366 t) and 4488 (304 t), respectively (Mejuto and Garc6s 1984; Mejuto 1985). Between April and August of 1986, the Spanish longliners spread westward toward the Grand Banks, where 423 makos were reported as a by-catch (Mejuto and Iglesias 1987). In subsequent years, the fishery expanded into the equatorial zones until, by 1990, the distribution of effort covered a large part of the Atlantic between 45'N and 05*S and east of 45'W. Most of the effort was concentrated in two areas: a southern region between 0 and 5'N off Africa, and a northern region between 35 and 45'N (Mejuto et al. 1992).

Despite a long history of angler interest in mako sharks in many parts of the world, the recreational fishery for sharks on the eastern coast of the United States essentially began off New York and New Jersey only about 30 years ago (Casey 1985). The estimated number of mako sharks caught by recreational anglers along the US Atlantic coast and in the Gulf of Mexico in 1978 was 17 973 (1223 t) (Casey and Hoey 1985). In recent years (1987-89), the US recreational catch from the Atlantic has been about 1000 t year-' (M. Holiday, Statistical Branch, National Marine Fisheries Service, personal communication).

In 1962, the US Bureau of Sport Fisheries and Wildlife initiated biological studies of sharks as part of a newly established programme in marine gamefish research at the Sandy Hook Laboratory in New Jersey. The programme continues under the National Marine Fisheries Service (NMFS) at the Narragansett Laboratory in Rhode Island and was described by Casey (1985). In brief, the research that is directed at studying the life history and biology of large Atlantic sharks includes a Cooperative Shark Tagging Program.

For the first several years of the tagging programme, NMFS biologists were responsible for tagging most of the sharks because there was little recreational or commercial fishing for sharks along the Atlantic coast in the 1960s. Moreover, tags that could be used effectively for large sharks with a minimum of risk to fishers had not yet been developed. As a result of the improvement in shark-tagging methods in the mid-1960s and of the increase in sportfishing for sharks during the 1970s and 1980s, anglers became the core of the tagging programme. Currently, there are approximately 4000 volunteer recreational and commercial fishers and biologists along the Atlantic coast of North America and Europe tagging sharks for the NMFS programme. From 1962 through 1989, approximately 84 000 sharks representing 47 species were tagged. This paper presents results from 2459 releases and 231 tag returns from mako sharks and offers a hypothesis to explain migration-distribution patterns of the mako shark in the western North Atlantic in relation to water temperature.

During the initial phase of the study, a variety of tag types was tested on sharks, with a fin tag (Jumbo Rototag) and a dart tag ('M' tag) chosen for use in the tagging programme. The 'M' tag is a small, harpoon-like tag that has been used almost exclusively on mako sharks during the past 20 years. The tag is composed of a stainless-steel dart head, a monofilament streamer, and a plastic capsule containing an instruction message printed in five languages. Descriptions of the tags and the tagging method are provided by Davies and Joubert (1967) and Casey (1985).

Mako sharks were tagged by cooperating fishers along the eastern coast of the United States and Canada (Fig. 1). The primary release area (87% of all makos tagged) was off the north-eastern United States between Cape Hatteras, North Carolina, and Cape Cod, Massachusetts (Fig. 2). In this area, as the water temperature increases in the spring, mako sharks move northward and inshore onto the continental shelf. Mako sharks were tagged primarily off Cape Hatteras from March to May, off New Jersey and New York from June to August, and off Georges Bank from August to October.

Recreational anglers represent the majority (51 %) of the 4000 participants in the tagging programme. This group includes professional charter- and party-boat captains and private boat owners. The sportfishing vessels range in size from small (<9 m) outboard-motor boats for inshore fishing to large (9-18 m) sportfishing cruisers that generally fish between 20 and 50 nautical miles offshore, though trips to 100 nautical miles are not unusual. Although mako sharks are sometimes caught by trolling, the principal sportfishing method involves casting ground or diced fish into the sea (chumming) to attract sharks to the vessel. The boat is allowed to drift (or is sometimes anchored), and baits are set at different depths in the 'chum slick' by means of floats (Barrett 1983). Once a shark is hooked, it is brought alongside the boat, a tag is inserted in the muscle near the first dorsal fin, the size (length or weight) of the shark is estimated, the sex is determined, and the leader is cut, leaving the hook in the jaw. Some fishers bring small sharks aboard and measure them; others estimate total or fork lengths from marks on their boats or divisions on their tagging poles. Occasionally, a shark is tagged while it is 'free swimming' near the boat or when it has been caught on a hand-line. Release information (including species, date, locality, size, sex, tackle, fish condition, and tagger's name and address) is recorded on a pre-addressed data card and mailed to the Narragansett Laboratory.

Commercial fishers make up a small proportion of all volunteer taggers (5%) but are important in tagging mako sharks (I I%) (Table 1). Participating commercial fishers are primarily US swordfish and tuna longline fishers who tag some of their incidental catches of sharks. US Atlantic longline fisheries

are to collect fisheries data and to ensure compliance with regulations. Between 1976 and 1988, Japanese longline vessels fishing for tunas in US waters provided opportunities to tag 944 mako sharks (Table 1). Japanese longliners have not fished within the US EEZ since 1988.

NMFS and other biologists' tagging efforts are conducted primarily from research vessels or opportunistically aboard recreational and commercial fishing vessels. Shark research cruises aboard NMFS vessels over the past 28 years have used pelagic longlines as the principal method of catching sharks in the area between Miami, Florida, and Georges Bank. The gear is essentially the same as the early US longline used for swordfish, but less gear is set (2 to 5 nautical miles) and it is retrieved within 4 h to reduce fishing mortality. The tagging procedures and information recorded are essentially the same as those for volunteer fishers except that more small sharks arc brought aboard and measured on research vessels.

Table 1. Releases of tagged mako sharks (L oxyrinchus) by area and by occupation of taggers

In this paper, all length data are expressed in terms of fork length (FL) and are defined as a straight-line measurement from the tip of the snout to the fork of the tail. When size at tagging was reported as total length (TQ or total weight (TW), these data were converted to fork length by means of the following relationships:

These relationships are based on length measurements to the nearest centimetre and total weights to the nearest 0-2 kg. All lengths used for measured length-frequency analysis were taken as fork lengths and not converted from any other size data. Length- frequency data were collected primarily from mako sharks landed at sportfishing tournaments held in the north-eastern United States during the months of June and July.

Distances between release and recapture localities are straight-line nautical miles. Tagging localities were grouped into the following areas: Grand Banks (35-50'N,35-60'W), waters off the north-eastern United States (35-50'N,60-80'W), waters off the south-eastern United States (15-35'N,60-82'W), and Gulf of Mexico (15-35'N,82-100'W) (Fig. 1).

From 1962 through 1989, 2459 mako sharks were tagged under the NMFS Cooperative Shark Tagging Program. Most of the sharks (91%) were tagged during the last 13 years. The increase in the number of releases is largely due to the efforts of Foreign Fisheries Observers and to increased participation by recreational fishers (Fig. 3). Moreover, despite the incentives to retain almost any size of mako shark for food or sale or as a trophy, fishers have been increasingly willing to tag and release makos in recent years. Shark tournaments along the north-eastern US coast have established minimum sizes, allow the landing of only one shark per boat per day, or have otherwise encouraged anglers to release sharks,

The I : I sex ratio for all measured mako sharks is the same as the ratio at birth reported by Stevens (1983). The sex ratio changes with increasing size, however. The I : I sex ratio holds true until approximately 240 cm, after which there is a shift to a preponderance of females (Fig. 5). This contrasts with length- frequency and sex-ratio data provided by Mejuto and Garc6s (1984) for mako sharks taken in the eastern Atlantic longline fishery. These data, for the area between Spain and the Azores, show a much higher percentage of males at sizes of more than 200 cm. Information on adults larger than 240 cm FL is extremely sparse, but a few males and females exceeding this size have been reported from both the eastern and the western Atlantic (Fig. 5) (Mejuto 1985).

In all, 231 tags (9-4% of releases) have been returned by fishers from 16 countries (Table 2). The tag-return rates according to the different capture methods at tagging are given in Table 3. The numbers of tags returned are roughly in proportion to the numbers of sharks released for each capture method, and they show that mako sharks survived capture by all methods.

The time at liberty ranged from I day to a maximum of 8-2 years, with 59% of all recaptures occurring within the first year and 95% within 4 years (Table 4). Age and growth studies of mako sharks in the western North Atlantic show that males and females grow at the same rate until they are about 4 - 5 years old and 230 cm FL. Females reach a larger size

Of the 231 sharks recaptured, 61 (26%) were caught within 100 n miles of the tagging locality and 148 (64%) were caught within 300 n miles. The remaining 83 recaptures were long-distance ones that exceeded 300 n miles (Fig. 6). The long-distance recaptures include 77 makos released off the north-eastern United States that were recaptured over an extensive area of the Atlantic, including the Caribbean and the Gulf of Mexico. Three returns from makos tagged in the Gulf of Mexico showed movements exceeding 300 n miles to Cuba, eastern Florida, and the Sargasso Sea. The two long-distance recaptures from releases on the Grand Banks and the single long-distance recapture from releases off the south-eastern United States occurred off the north-eastern United States. The maximum straight-line distance travelled by a tagged mako shark was 2452 n miles between the north-eastern United States and Spain. This 160-cm-FL individual was tagged in February 1984 on the southern edge of Georges Bank by a Foreign Fisheries Observer on a Japanese tuna vessel and was recaptured by a Spanish longliner in November 1984.

The long-distance recaptures do not show trends that can be related to size or sex for the sizes represented in the data. Makos ranging in size from 69 to 239 cm showed movements exceeding 500 n miles, with the 69-cm individual travelling 588 n miles in 120 days. Although the distribution and abundance of mako sharks vary with the season in the northern parts of the sharks' range, the long-distance recaptures do not show a well defined seasonal pattern. The 31 recaptures that exceeded 1000 n miles occurred in all months of the year (Table 5).

In the North Atlantic, the preferred water temperature of the mako shark appears to lie in a rather narrow range between 17 and 22'C. Hoey (1983) analysed surface-temperature data from 2766 sets of swordfish longline gear between the Gulf of Mexico and the Grand Banks and found that the mean minimum and maximum temperatures in which makos were caught were 18-5 and 20-5'C, respectively. The Spanish swordfish longline fishery which covers an extensive area between Spain and the Grand Banks, also shows the highest catch rates for mako sharks in surface temperatures of 18'C in the region east of the Grand Banks (41-47'N,38-40'W) (Mejuto and Iglesias 1987). Hoey (1983) reported that the overall surface-temperature range in which fishing gear was used was 4-3 PC, which was largely due to pelagic longline gear being set along strong surface-temperature anomalies, particularly along the margins of the Gulf Stream north of Cape Hatteras and in the vicinity of the Grand Banks. Other than the blue shark (Prionace glauca), the mako shark was the most common shark taken in the swordfish longline fishery north of Cape Hatteras. Blacktip sharks (Carcharhinus sp.) dominated the by-catch in the swordfish fishery in the Gulf of Mexico and off the south-eastern United States in a preferred temperature range of 22-27C (Hoey 1983).

Recreational catches of mako sharks off the middle Atlantic states of the United States also show that mako sharks do not move inshore onto the continental shelf in spring until surface temperatures reach about 17C. The oldest continuous mako shark-fishing tournament in this region has been held at Bay Shore, New York, since 1965 (Casey et al. 1978). The trend in the species-composition data from this tournament, which is held in late June, shows that blue sharks dominated catches when surface temperatures were below 18'C but that mako sharks were common, and in some recent years dominant, when water temperatures reached 18'C.

Other evidence on the water temperatures preferred by mako sharks is provided by sonic tagging experiments. In one experiment, a temperature-monitoring radio tag was attached to a 160-180-kg mako shark that was followed for 4 days on a track from Florida across the Gulf Stream and into the Sargasso Sea, north of the Bahamas (Carey et al. 1978). The shark swam at depths ranging from the surface to 500 m but spent most of its time at

Since the mako shark is warm-bodied (Carey and Teal 1969) and is widely distributed throughout the world's oceans, individual makos are probably able to withstand temperatures outside the 'normal range' suggested here. Nevertheless, the evidence that makos occupy a depth range from the surface to at least 500 m and prefer water temperatures near I8'C allows us to consider some basic ideas to explain the pattern of tag returns from the western North Atlantic (Fig. 6).

Worthington (1959) described the source of 18'C water in the Sargasso Sea as northern Gulf Stream water that, in winter, cools from 20'C and sinks to form an isothermal layer that flows south beneath the warmer waters of the Sargasso Sea. In the northern Sargasso Sea (north of 32'N), this layer is never less than 250 m thick and it reaches its greatest depth of about 400 m in early March. Worthington (1959) describes the 18'C layer as a wedge 250 m thick, extending 12 000 km along the northern margin of the Gulf Stream (roughly between 75 and 45'W) and extending 1500 km into the Sargasso Sea. The eastern edge of the 18'C water is not well defined but is approximately over the Mid-Atlantic Ridge.

Keasons other than temperature can be advanced to explain the distribution of mako recaptures. The data presented here are based largely on tag returns derived from the by-catches of fishing fleets. Collectively, the longline fisheries discussed previously together with those of other countries, extend over broad regions of the North Atlantic and represent potential sources of mako shark tag returns. However, by-catches of mako sharks are influenced by the seasonal patterns of traditional fishing grounds, by gear characteristics, including line material, hook type and fishing depth, and by biological factors such as the distribution, abundance and behaviour of different target species. The paucity of tag returns from the eastern Atlantic may be due to the relatively low overall number of makos that have been tagged compared with other species. For example, approximately 6000 blue sharks were tagged in the western North Atlantic between 1962 and 1974 before one was recaptured east of the Mid-Atlantic Ridge. Finally, under-reporting of tags is a factor that exists, but it is difficult to assess. Although the instructions on the tag legend are printed in English, Spanish, Norwegian, French and Japanese, some fishers simply disregard tags. Other fishers may not return tags because they feel it is not in their political or personal interests to provide information on their fishing activities. Despite all of these shortcomings, tag and recapture data offer some insight into the distribution and movements of mako sharks in the western North Atlantic.

Mako sharks are widely distributed throughout the western North Atlantic in a preferred temperature range of 17-22'C. A seasonal cycle of abundance off the north-eastern United States, beginning in January, shows that L oxyrinchus is common along the western margin of the Gulf Stream, with at least one area of high abundance (based on commercial catches) off Cape Hatteras, where the Gulf Stream flows near the continental shelf. From January through April, makos are seldom taken on the continental shelf north of Cape Hatteras. Beginning in April and May, as inshore shelf waters warm and the axis of the Gulf Stream moves farther north, mako sharks begin moving northward onto the continental shelf between Cape Hatteras and the southern part of Georges Bank. In most years, makos are taken off southern New Jersey in early June and off New York and the southern New England states by late June. The movement inshore and northward takes place simultaneously along the Gulf Stream from Cape Hatteras to the Grand Banks (Hoey 1983; Mejuto and Iglesias 1987).

From June through October, makos are caught by anglers on the continental shelf between Cape Hatteras and Cape Cod and, in recent years, in southern parts of the Gulf of Maine. During this season, swordfish and tuna longline fishers also catch makos between the continental shelf and the Gulf Stream from Cape Hatteras to the southern tip of the Grand Banks. This entire area, particularly on the continental shelf south of Cape Cod, may be the primary feeding grounds for a large part of the juvenile and subadult mako population in the western North Atlantic. Off the north-eastern United States, 33 different food items were found in mako stomachs, with bluefish (Pomatomus saltatrix) representing 78% of the prey by volume. The percentage of mako sharks with food in their stomachs and caught inshore between Cape Hatteras and Cape Cod was significantly higher than the percentage of such sharks caught offshore beyond the continental shelf (Stillwell and Kohler 1982).

During late fall and early winter (November-December), makos move from the area between Cape Hatteras and the Grand Banks to offshore wintering grounds in the Gulf Stream and the Sargasso Sea. If we assume that 18'C Sargasso Sea water represents the preferred habitat for makos, then the core of their distribution in the western North Atlantic covers a latitudinal range between 20 and 40'N, bordered by the Mid-Atlantic Ridge on the east and the Gulf Stream on the west. Certainly, mako sharks occur outside of these boundaries at different seasons, and they can make transatlantic crossings as evidenced by

the single recapture from the coast of Spain (Fig. 6). However, most of the recaptures can be explained on the basis of a 'Sargasso Sea' hypothesis, including those returns from the Caribbean Sea. Worthington (1959) reports that some northern Sargasso Sea water enters the Caribbean, presumably through the Windward Passage; his studies and those of Fuglister (1960) show an 18'C temperature and 36-5 salinity layer between 125 and 300 m depth in the Caribbean. The route that makos might travel in this water into the Caribbean and then into the Gulf of Mexico and the Florida Straits is consistent with the tag returns from those areas. However, the distribution of recaptures suggests that the principal wintering grounds of juvenile makos are the western margin of the Gulf Stream and the northern part of the Sargasso Sea.

If the above hypothesis offers a partial explanation for the pattern of tag returns for the mako shark in the western North Atlantic, the picture is far from complete. Moreover, attempts to understand the life history of L oxyrinchus cannot truly advance until more is known about the reproductive biology of the species and the distribution of large adults. Only four pregnant specimens of L oxyrinchus have been reported from the western North Atlantic (Depperman 1953; Guitart Manday 1975; Branstetter 1981; R. G. Gilmore, Harbor Branch Inst., Ft Pierce, FL, personal communication); all were taken in southern areas off Florida and in the Gulf of Mexico. On the other hand, very small makos, considered to be recently born, have been reported during the summer months in an area extending from the Gulf of Mexico to the Grand Banks. One explanation for this distribution of young is that adult females remain far offshore and/or are widely distributed in tropical waters. If so, newborn makos may be dispersed over a broad geographical area by the Gulf Stream. A reproductive strategy that ensures widely separated pregnant females and the dispersal of young over a broad area would offer some protection for the young from predation by large oceanic predators, including other mako sharks.

Results of 231 tag returns from mako sharks tagged in the western North Atlantic suggest that makos are concentrated within a principal water temperature range of 17-22'C over a geographical range extending from the Mid-Atlantic Ridge to the western margins of the Gulf Stream, the Caribbean and the Gulf of Mexico. Tag returns show only one transatlantic movement by a mako shark and suggest a separate stock in the western North Atlantic. However, any conclusions regarding stock relationships must be considered preliminary in light of an alternative hypothesis recently proposed for swordfish.

An analogous situation exists for tagging studies on swordfish: despite nearly 2000 swordfish tagged in the western North Atlantic, none have been recaptured in the eastern Atlantic. Nevertheless, Mejuto and Hoey (1991) present evidence in support of a continuous distribution of the swordfish stock for the entire Atlantic. Their hypothesis, based on analysis of several years of longline catch data, comparison of size distributions, maturity, and egg and larvae distributions, proposes north-south movements of swordfish from common spawning grounds in the tropical and equatorial zones of the Atlantic. The principal reasons given for the absence of transatlantic tag returns from swordfish are the low number of releases to date and the uneven distribution of fishing effort (Mejuto and Hoey 1991). A similar interpretation could be applied to tag-return data for mako sharks.

Unquestionably, additional studies are required to significantly advance knowledge of the biology of L oxyrinchus and to provide information for future management of the stocks. The importance of the mako shark in US fisheries is clearly recognized in the NMFS Fishery Management Plan for Sharks of the Atlantic Ocean (Anon. 1991). The mako is the only species in this plan covering 39 shark species for which minimum size regulations are proposed. For international management, new initiatives should be directed toward: (1) the collection of fisheries data by all countries that utilize or incidentally catch mako sharks,

We are indebted to the hundreds of volunteer anglers, commercial fishers, Foreign Fisheries Observers and biologists who made this research possible by tagging their catches. We are particularly grateful to longline Captains Philip Ruhle and Philip Ruhle Jr of Newport, Rhode Island, to Foreign Fisheries Observers Georg Hinteregger and Jerzy Cygler, and to Captain Stephen Connett of the research vessel Geronimo, all of whom tagged many mako sharks. The staff of the Apex Predator Investigation deserve thanks, especially our colleagues Harold (Wes) Pratt, Charles Stillwell, Gregory Skomal and Lisa Natanson, who spent long hours on research cruises tagging sharks, and Patricia Turner, who assisted in preparing the charts and graphs. We appreciate the work and counsel of Francis Carey, who has always stimulated our thinking about sharks. We also thank Jaime Mejuto, John Hoey, Ray Conser and Kenneth Sherman for reviewing the manuscript and providing useful comments and suggestions.

Barrett, P. (1983). 'Fishing for Sharks.' (Barrett Productions: Brick, New Jersey.) 132 pp.

Casey, J. G. (1985). Transatlantic migrations of the blue shark: a case history of cooperative shark tagging. In 'World Angling Resources and Challenges: Proceedings of the First World Angling Conference, Cap d'Agde, France, September 12-18, 1984'. (Ed. R. H. Strond.) pp. 253-68. (International Game Fish Association: Ft. Lauderdale, Florida.)

Casey, J. G., Mather, F., 111, Mason, J., and Hoenig, J. (1978). Offshore fisheries of the Middle Atlantic Bight. In 'Proceedings of the Second Annual Marine Recreation Fisheries Symposium, Norfolk, Virginia, March 29-30, 1978'. (Ed. H. Clepper.) pp. 107-29. (International Game Fish Association, NCMC, Sport Fishing Institute: Washington, DC.)

Guitart Manday, D. (1975). Las Pesquerias Peldgico-Ocednicas de Corto Radio de Acci6n en la Regi6n Noroccidental de Cuba. [Short-range marine pelagic fishing of northwest Cuba.) Seria Oceanologica, Oceanographic Institute, Academy of Sciences of Cuba. No. 31, pp. 3-26. [Translated for the National Science Foundation and the US Department of Commerce, National Oceanic and Atmospheric Administration, National Marine Fisheries Service, Washington, DC, by the Agence Tunisienne de Public-Relations, Tunis.]

Hoenig, J. M., and Gruber, S. H. (1990). Life-history patterns in the elasmobranchs: implications for fisheries management. Elasmobranchs as living resources: advances in the biology, ecology, systematics, and the status of the fisheries. (Eds H. L. Pratt Jr, S. H. Gruber and T. Taniuchi.) NOAA Technical Report NMFS No. 90, pp. 1-16.

Mejuto, J., Sanchez, P., and de la Serna, J. M. (1992). Nominal catch per unit of effort by length groups and areas of the longline Spanish fleet targeting swordfish (Xiphias gladius) in the Atlantic, years 1988 to 1990, combined. International Commission for the Conservation of Atlantic Tunas SCRS 91/49.

Worthington, L. V. (1959). The 18' water in the Sargasso Sea. Deep-sea Research 5, 297-305.