Florida's Halfbeak, Hemiramphus spp., Bait Fishery

RICHARD MCBRIDE, LISA FOUSHEE, and BEHZAD MAHMOUDI

Introduction cies is in southeastern Florida. Neither of flyingfishes (Exocoetidae) and species is common north of Florida needlefishes (Belonidae) are consis­ In Florida, ballyhoo, Hemiramphus (Hildebrand and Schroeder, 1928; tently caught as bycatch (Berkeley et aI., brasiliensis, and balao, H. balao, form Hardy and Johnson, 1974; Hardy, 1978; 1975). the basis of an important bait fishery Burgess et aI., 1979), and only nominal This halfbeak fishery dates back to (Berkeley et aI., 1975; Berkeley and landings are reported elsewhere, in the 1950's (Siebenaler, 1955), and an­ Houde, 1978). They are harvested with South America (Collette, 1978), Puerto nual landings increased rapidly in the modified lampara nets, a type of sur­ Rico (Kimmel, 1991), and the Virgin late 1960's because demand for bait rounding net (Nedelic, 1982), and are Islands (Beets and La Place, 1991). from marine recreational anglers had sold as bait to anglers seeldng pelagic Ballyhoo and balao are both known increased (Berkeley et aI., 1975). Also and benthic gamefishes (e.g., billfishes as halfbeaks (Hemiramphidae; Collette during this time, the fishery increased and groupers). In the western Atlantic, et aI., 1984), and because they are its fishing power; this was accomplished both species are found from at least New equally acceptable as bait, the fishery by using larger nets with long ropes at­ York to Brazil (Collette, 1965), but the does not sort the catch by species (Ber­ tached at the ends to herd the fish. This only significant fishery for either spe- keley et aI., 1975). Still, they can be rapid increase in landings during the late readily distinguished by the length of 1960's was a reason for the investiga­ The authors are with the Florida Marine Research the pectoral fin and by their coloration tions by Berkeley et ai. (1975) and Ber­ Institute, Florida Department of Environmental Protection, 100 Eighth Avenue S.E., SI. Peters­ when alive (Fig. 1; Berkeley et aI., 1975; keley and Houde (1978). They con­ burg, FL 33701-5095. Telephone 813-896-8626; Collette, 1978; Bohlke and Chaplin, cluded that overfishing was not occur­ Fax 813-823-0166; Corresponding email: [email protected]. All photo­ 1993). Both halfbeak species frequently ring during the early 1970's, but no as­ graphs are by and courtesy of the senior author, occur together in individual fishing sessment of the commercial fishery has Richard McBride. hauls, but ballyhoo constitutes the ma­ been made since. jority of the catch during summer and In this paper we describe landings, ABSTRACT-Two species of haljbeaks, nearly all of the catch during winter markets, and fishing methods for ballyhoo, Hemiramphus brasiliensis, and balao, H. balao, form the basis of a rela­ (Berkeley et aI., 1975). Mixed samples halfbeaks in Florida. We also assess the tively small but valuable bait fishery in from throughout the year contained only current state of the fishery based on spe­ southeastern Florida. Haljbeak landings 9.8% balao (Berkeley and Houde, cies composition and on measurements increased rapidly in the late 1960's but are 1978). Although other species also ofcatch, fishing effort, and size frequen­ now relatively stable (about 450,000 kg or Hyporhamphus 1 million lb annually), and their ex-vessel known as halfbeaks (e.g. cies. The data available for analysis are price is about $600,000. Fishing methods, spp.) are suitable for bait, they are not part historical landings for 1950-94, as well which had changed in the late 1960's when of the fishery because they occupy dif­ as onboard observations and dockside landings increased, have changed little ferent habitats and are not found in con­ interviews completed during 1988-91. since the 1970's. Data from a fishery-de­ centrated schools (Tabb and Manning, These data are compared with histori­ pendent survey (1988-91) show that catch 1961; Banford and Collette, 1993; cal data reported by Berkeley et ai. rates were highest from October to Febru­ 1 ary, when catches were dominated by large Reintjes ). Florida's halfbeak fishery is (1975) and, in particular, are used to ballyhoo (>200 mm or 8 inchesfork length highly selective, and only low numbers evaluate their prediction that exploita­ (FL)); rates were lowest from May to Sep­ tion of halfbeaks would increase in the tember, when catches contained both spe­ future because of an expanding recre­ cies in more equal numbers and the size range was greater (about 150-250 mm FL) lReintjes, J. W. 1979. Coastal herrings and asso­ ational fishery in Florida. than it was for winter landings. There was ciated species: a profile of species or groups of little bycatch, and only flyingfishes species, their biology, ecology, current exploita­ Materials and Methods tion with economic and social information. Rep. (Exocoetidae) and needlefishes (Belonidae) prep. for GulfMex. Fish. Manage. Counc., South­ Commercial landings have been occurred consistently. Comparisons of the east Fish. Sci. Cent., NMFS, NOAA, Beaufort, monitored in Florida since 1950 by the 1988-91 data with similar data reported N.C., 170 p. All footnoted reports are on file at from 1974 indicated that haljbeak popula­ the Florida Marine Research Institute, 100 Eighth State of Florida in cooperation with the tions have remained relatively stable. Avenue S.E., St. Petersburg, FL 33701-5095. U.S. Fish and Wildlife Service or the

58(1-2), 1996 29 Above: Striking the net lines during haulback to keep the surface-oriented halfbeaks from escaping.

Above: Selling the lampara net with a marker buoy off Key Largo. When all the net and ropes are paid out the area swept can be more than a square mile.

Right: Close-up of halfbeaks in the purse of the lampara net.

30 Marine Fisheries Review Left: Halfbeaks in the purse of the lampara net. A needlefish is being re­ moved from the net purse and released alive.

Below: Halfbeacks are immediately layered in ice and brine to pre­ serve their condition for trolling bait.

Above: Halfbeaks are sorted by size and condition, not by species, at the fish house. The largest sizes are rigged with one or two hooks and a wire leader before being flash frozen.

58(1-2), 1996 31 Figure I.-Hemiramphus batao, balao (top); when alive the upper lobe of the caudal fin is bluish-violet with a red tip and the lower lobe is bluish. Hemiramphus brasiliensis, ballyhoo (bottom); when alive the upper lobe of the caudal fish is yellowish orange and the lower lobe is dusky. Illustrations provided by Bruce Collette, NMFS National Systematics Laboratory.

National Marine Fisheries Service. County, and Monroe County (Fig. 2). Results and Discussion From 1950 to 1985,Iarge foodfish pro­ The survey monitored 1) species com­ Landings and Markets cessors were the sources of these data. position and abundance of target and In 1985, a new survey design, the Ma­ bycatch species based on subsamples of Annual halfbeak landings have pro­ rine Fisheries Information System 3-11 kg of a set's catch or up to 23 kg gressed through three phases during the (MFIS), was implemented by the Florida of a day's catch; 2) fishing effort (i.e., past four decades: 1) a phase when fish­ Department of Environmental Protection frequency and duration of fishing days; ing rates and demand were low from (FDEP) (Muller and Murphy2). Florida's frequency and duration of individual 1950 to about 1965, 2) a phase during MFIS requires that records be kept ofev­ hauls for onboard observations); and 3) which the annual landings rapidly in­ ery wholesale transaction of marine or­ length-frequency of ballyhoo and balao creased-1966-70, and 3) a phase when ganisms landed in Florida, so estimates based on subsamples (100-300 fish per landings have been relatively high and of baitfish landings since 1985 are more fishing day) of the catch. Fork lengths stable since about 1971 (Fig. 3). The accurate than previous estimates were. (FL) were measured to the nearest mil­ threefold to fourfold increase in land­ From November 1988 to August limeter from the tip of the upper jaw to ings during the late 1960's was attrib­ 1991, the FDEP also completed a sur­ the fork of the tail. Individual fish from uted by Berkeley et aI. (1975) to increas­ vey of the halfbeak fishery independent some subsamples were also weighed to ing demand for bait as well as to the of the MFIS program. We made on­ the nearest 0.1 g to establish length­ use of larger and more efficient nets. board observations (33 vessel days) and weight relationships. Catch rates were Another threefold increase between dock-side interviews (an additional 70 calculated as: 1985 and 1986 was largely an artifact vessel days), which covered 26 of the of a change in data collection methods 34 months in the survey period. Samples associated with the implementation of CPUE= CIf, (1) were obtained in Florida's southeastern the Florida's MFIS program. counties, where the state's halfbeak Since landings prior to 1986 were landings are greatest, and sampling ef­ where CPUE = catch per unit ofeffort, only from large foodfish processors, fort was apportioned between three sub­ C = catch in numbers or there was great potential to underesti­ regions: Martin County to Palm Beach weight of fish mate landings of halfbeaks and other County, Broward County to Dade f one vessel-day (one baitfish species (Broadhead, 1951; day of fishing by one Reintjes 1; Johnson3). Annual landings vessel). during 1976-81 averaged 164,000 ± 2Muller, R. G., and M. D. Murphy. 1994. Report on inshore finfish trends. Rep. prep. for Fla. Mar. Catch is reported by weight unless Fish. Comm., Dep. Environ. Prot., Fla. Mar. Res. otherwise noted. Notes regarding fish­ 3Johnson, L. E. 1974. Commercial fisheries of Inst., St. Petersburg, June 13, 1994, 16 p. Prior Florida, 1973. In Summary of Florida commer­ to 1994 the FDEP was known as the Florida De­ ery markets and fishing methods were cial marine landings, 1973. Fla. Dep. Nat. partment of Natural Resources. also recorded during 1988-91. Resour., St. Petersburg, 62 p.

32 Marine Fisheries Review 40,000 kg (360,000 ± 87,000 Ib; mean ± standard deviation) compared to Atlantic 477,000 ± 86,000 kg (1,049,000 ± Ocean 189,600 Ib) during 1986-91. The change in reporting methods had the effect of tripling the landings. Based on the assumption that ha1fbeak landings have been relatively stable since 1971, then landings during this third and most recent phase have been maintained at Gulf of Mexico about 480,000 kg (1.05 million Ib) per year. Since 1950, halfbeaks have been landed almost exclusively in southeast­ ern counties of Florida (Fig. 3,4). Ef­ fort was also greatest in southeastern counties, principally in Palm Beach, Dade, and Monroe counties (Fig. 4). Notable landings and effort also oc­ curred in the Florida panhandle, but this constituted bycatch from other bait fish­ eries there. The markets have changed little since being described by Siebenaler (1955) and Berkeley et al. (1975). Halfbeaks Figure 2.-Florida's coastal counties and areas of major fishing activity for halfbeaks sur­ veyed in 1988-91. Counties mentioned in the text or Figure 3 are identified; Monroe County are frequently the preferred bait for includes all of the Florida Keys. Resolution for identifying fishing zones while at sea was 5' marine anglers seeking billfishes lat. x 5' long. (Istiophoridae), dolphin, Coryphaena hippurus, groupers (Serranidae), and 1.50--,------, other gamefishes (de Sylva, 1974; o Gulf of Mexico coastal counties Nakamura and Rivas, 1974; Gentle, • Atlantic Ocean coastal counties 1977; Jolly, 1977). The Gulf Stream is 1.25 very close to Florida's southeast coast, so offshore fishing for pelagic .c 0.50 gamefishes is very accessible. Whole­ - Change in Cl o 1.00 -~ sale fish houses sell halfbeaks to char­ o Statistical o 0 ter boat operators, tackle shops, and in­ Reporting 0 o 0 dividual anglers primarily within o 0 o~ 0.75 ~ 0 Florida, but markets also exist out of .... 0 state and abroad. Small halfbeaks are -en .... Cl used for drift fishing for mackerels and c::: 0.25 ­en snappers, whereas larger sizes are used :c 0.50 Cl c::: c::: more commonly for trolling for off­ tIS '0 shore, pelagic gamefishes. In the Pa­ ...J c::: tIS cific, halfbeaks (Arrhamphus, Hemi­ 0.25 ...J ramphus, and Hyporhamphus) are sold as both bait and food (Collette, 1974), but to our knowledge halfbeaks landed 0.00 0.00 in Florida are not sold for food. 50 60 70 80 90 Ex-vessel prices (e.g. those paid di­ Year rectly to fishermen) varied principally in relation to fish size-class and seasonal Figure 3.-Florida's annual landings of halfbeaks (ballyhoo and balao combined), 1950­ availability. Halfbeaks were usually 1994. Data for the years 1950-85 were collected from commercial fish processors by the State of Florida. Data for 1986-94 were based on Florida's MFIS. Atlantic coastal counties sold by the piece (instead of by weight), include Monroe and all others east; Gulf coastal counties include Collier and all others to the which differs from marketing of other west (Fig. 2). Landings for Gulf coast counties equal zero for 16 of 45 years, but Gulf coast landings are frequently not visible here unless greater than or equal to 5,000 lb because of the baitfishes. Conversions between num­ range of the ordinate (left axis=lb, right axis=kg). The annual landing for 1964,917,000 bers and weight in this study used an lb(*), is anomalous and cannot be explained.

58(1-2}, 1996 33 average value of 13.7 fish/kg (6.2 (127-156 g or 4.5-5.5 oz), horse (156­ halfbeaks as thread herring, Opistho­ halfbeaks/lb), which was based on un­ 184 g or 5.5-6.5 oz), and jumbo (184+ nema oglinum, were landed, and only sorted samples from the fishery col­ g or 6.5+ oz). The ex-vessel prices half as many halfbeaks as Spanish sar­ lected year-round (n =3524). Some of ranged from $0.02-0.03/fish for small dines, Sardinella aurita, were landed in each day's catch was sold as fresh bait, sizes to $0.12-0.20/fish for horse or 1989-90 (comparisons done by which was not sorted by size and con­ jumbo sizes. In addition, some very weight2). However, the 1989-90 value sistently commanded a high price of small (about 1oz) or damaged fish were of these two clupeid species was only $0.15/fish (about $2.00/kg or $0.90/lb) typically sold at a bulk price, which was $0.22-0.33/kg ($0.1O-0.15/lb), so the during 1989-90. Except during busy consistently $0.33/kg ($0. 15/lb). The annual ex-vessel value of Atlantic sportfishing seasons, when demand for average value for halfbeaks landed in thread herring was only about $500,000 fresh bait was high, a majority of 1989 was $1.43/kg ($0.65/lb) based on and Spanish sardine was only about halfbeaks were sorted by size and then ex-vessel transactions in Palm Beach $350,000. frozen. Representative size categories4 and Dade counties. The 1990 price was Fishing Methods were small (85 g or <3 ounce [oz]), estimated as $1.l9/kg ($0.54/1b), but medium (85-127 g or 3-4.5 oz,), large this was not significantly different be­ During 1988-91, ten vessels were tween years based on 95% confidence operating in southeast Florida, each intervals of $1.06--1.80/kg ($0.48-0.82/ with a captain and two or three crew Ib) in 1989 and $0.79-1.58/kg ($0.36­ members. These boats ranged from 7.3 O.72/lb) in 1990. m to 10.4 m (24-34 ft) long and were A) Regional landings Ex-vessel prices during 1989-90 powered by 250-350 horsepower en­ 2,900,000 kg from 1986-1991 were considerably higher than values gines. Their holding capacities were reported by Berkeley et al. (1975) but limited to six-eight 0.45 m3 (16-ft3) were similar after adjustment for infla­ boxes; each box could hold about 230 tion. In 1974, ex-vessel prices were kg (500 lb) or about 2,000-3,000 indi­ $0.03-0.08/fish compared to the aver­ vidual fish. Fish were captured with age range of $0.03-0.15/fish in 1989­ modified lampara nets, a type of sur­ 90. Retail prices in 1974 were $1.75­ rounding net (Nedelec, 1982). The nets 3.00/dozen ($0. 15-Q.25/fish) compared measured 410-640 m (450-750 yd) in to $4.50-5.00/dozen (0.38-0.42/fish length and had wings that hung 1.2-2.4 and as high as $1.00/fish for the larger m (4-8 ft) deep; the wings are cut to sizes) in 1989-90. Inflation alone would fish only the upper water column be­ cause an ex-vessel price of $0.05 in cause halfbeaks are found near the sur­ 1974 to rise to $0.12 in 1989,5 whereas face in association with reef habitat. A the retail value would rise from $0.20 pocket 10.0- to 12.1-m (33-40 ft) deep in 1974 to $0.48 in 1989. was sewn in at one end of the net. The Compared with Florida's other mesh used ranged from 1.9 cm to 3.8 baitfish species, halfbeak landings were cm (0.75-1.5 in) stretch mesh. These relatively small but the value was quite nets were modified from a typical high. In 1989-90, annual ex-vessel lampara design, by having purse lines value of the commercial fishery was in the pocket. An additional 550-3,660 about $600,000, based on landings of m (600-4,000 yd) of line was attached 0.477 million kg (1.05 million Ib) mul­ to the net to increase the area swept tiplied by the estimated prices above. when fishing the net. Relative to other important baitfishes Ballyhoo and balao school in coastal in Florida, only one-fourth as many waters near and between reef structures (Tabb and Manning, 1961; Stark and Davis, 1966), where they can be ob­ 4Up to seven categories could be used at differ­ served near the surface. An average day ent fish houses. Berkeley et al. (1975) reported four categories: <185 mm, 185-235 mm, 235­ of fishing on the vessels surveyed oc­ B) Regional fishing trips 265 mm, and >265 mm fork length. Those cat­ curred during daylight and lasted 8,410 trips from 1986-1991 egories are similar to these reported here and slightly more than seven hours. Shal­ correspond approximately to <50 g, 50--110 g, 110--160 g, and >160 g, based on length-weight low inshore waters were searched out Figure 4.-Regional halfbeak (ballyhoo conversions from Berkeley and Houde (1978). to 9 km (5 n.mi.) from shore and to a and balao combined) landings (A) and SInflation is adjusted for a price difference of2.42 depth of about 18 m (60 ft). The pri­ number of fishing trips (B) by county or between 1974 and 1989, using the Economic mary fishing grounds identified during groups of counties. See Figure 2 for Report of the President, January J993, Table b­ county locations. Data are for 1986-1991 3: Implicit price deflator for gross domestic prod­ our survey were the nearshore patch from Florida's MFIS. uct,1959-1992. reefs in the Palm Beach and Biscayne

34 Marine Fisheries Review Bay regions and the offsh?re reefs of Species Composition ranged from 360 to 1,720 kgld (800­ the Biscayne Bay and Flonda Keys re­ The percentage composition of bal­ 3,800 Ibid), but balao catches rarely.ex­ gions (Fig. 2). Fishing depths were lyhoo and balao changed seasonally ceeded 450 kgld (1,000 Ibid) at any tIme (abo~t greater in Palm Beach County (Fig. 5), similar to the changes reported of the year. These seasonal variations 9-15 m or 30-50 ft), where the contI­ by Berkeley et al. (1975). Pure catches in catch rates appear to be affected by nental shelf is narrower, than in Dade both the absolute abundance and the of ballyhoo occurred (45 of 89. obs~rved or Monroe counties «9 m or 30 ft). availability of fish. Winter catches were hauls, 50.5%), principally In Winter More area was searched during April­ higher in part because age-O fish grow months. Ballyhoo and balao occ~rre.d September when fish were relatively together in 35 hauls (39.3%), pnncI­ fast enough to become exploited (see scarce, but in general only one or tw.o below). Fish behavior is also a factor, pally during spring-autumn. P~re 5' x 5' (minutes of latitude and longI­ catches of balao occurred only tWIce because halfbeaks are more aggregated tude) grids were searched during a (2.2%). The remaining seven observed and in larger schools during winter on single day. The grids searched were far­ hauls (7.8%) caught neither species be­ these fishing grounds. Interspecific dif­ ther south during the winter and more cause of entangled nets, difficult ferences in abundance are not under­ to the north during summer, which pre­ weather conditions, or fish becoming stood but we are now (i.e. 1996) col­ sumably followed the seasonal move­ spooked and escaping. lecting data that may clarify the extent ments of the halfbeaks. The fishing gear and methods used Once a school was located, the net were highly selective for fish in the up­ and lines were set to surround it. The per water column, so bycatch was gen­ lines skimmed the water surface as the erally low. The only species frequently o net was drawn in, and less line was re­ <1l collected besides halfbeaks were quired when the fish were more closely .J:J flyingfishes (Exocoetidae: Cypselurus '" aggregated. Fish were loa?ed on~o the ! 75 and Parexocoetus) and needlefishes o boat with a dip net and Immediately o (Belonidae: Ablennes and Tylosaurus). .c placed in layers of crushed ice and 2­ Flyingfishes occurred most freq~ently brine. Setting and hauling lasted about .J:J during spring and summer (maxImum '" one hour, and three hauls were made per '0 of 16.6%, by numbers, in any observed C1l fishing day on average. The mean num­ g' 50 haul). Flyingfishes could be sold as b~it C ber of hauls made per day was lowest C1l and were therefore not always dIS­ ~ in summer and highest in autumn (Janu­ C1l 1974 carded. Needlefishes appeared sporadi­ ll. ary to March = 2.9, April to June = 3.0, 1988-91 cally (maximum of 5.1 %, by numb~rs, July to September 2.3, and October = in any observed haul). Other species to December = 3.5, based on onboard occurred only rarely and did not include 25 1 2 3 4 5 6 7 8 9 10 11 12 observations only). gamefish or foodfish species. Month Only subtle changes in fishing meth­ Figure 5.-Percentage composition of ods have been made recently. The size, Abundance halfbeaks, ballyhoo vs. balao by month, storage capacity, and horsepower of the in the southeast Florida lampara net fish­ Catch rates of ballyhoo peaked dur­ vessels observed by Berkeley et al. ery. Data source is based on onboard ob­ ing winter, whereas catch rates ofbalao servations of 8,504 fish from 80 different (1975) were slightly less than those peaked during summer (Fig. 6). Bally­ hauls made on 32 days, 1988-1991. observed during 1988-91. Larger ves­ Background data for 1974 is from Ber­ hoo catches during October-March sels can hold more fish, but because keley et al. (1975). b~it, many fish are sold as fresh da!ly 4.000 catch is closely associated wIth dally >­ >­ '9'" Ballyhoo demand at the fish houses and bait Qi -0­ Balao 1,500 '9'" 3.000 - Qi shops. The nylon nets used during "' "' "'>'" "' Cl '"> 1988-91 were similar to those reported l: 1,000 Cl :E 2,000 l: by Berkeley et al. (1975) but were much ~ :E .... ~ larger than the cotton nets descri~ed by .... ~ 1,000 500 C;; Siebenaler (1955). Nets used In the J: ~ J: early 1970's were 370--600 m (400--650 16" 0 B yd) long, whereas nets used in the 0 0 0'" 1950's were only 180-275 m (200-300 Jan. May Sept. Jan. May Sept. Jan. May Sept. yd) long. The use of extra line to herd 1989 1990 1991 the fish during the fishing operation also Figure 6.-Mean daily catch/vessel-day for.26 months s~~pled began in the 1960's. Thus, th~ m.ost from November 1988 to August 1991 (left aXls=lb, nght aXis-kg). Data depicts catch rates of ballyhoo and balao separately; dashed notable changes in halfbeak fIshing lines indicate sampling across nonconsecutIve months. Error bars technology occurred during the 1960's. are ± I s.e.

58(1-2), 1996 35 of habitat specificity and seasonal migra­ casions are rare, however, and we ob­ inches) FL at annulus I (Berkeley and tion patterns by each halfbeak species. served this only once when an estimated Houde, 1978). Berkeley and Houde also Catch rates for both halfbeak species 1.5 boxes of fish were released live af­ compared growth rates from the labo­ combined declined in summer, which ter 3 boxes were packed during a March ratory and field and concluded that the also was the seasonal trend observed in fishing trip. size-frequency of age-O fish in the fish­ 1974 (Fig. 7). Monthly catch rates dur­ ery is probably biased until October or Size Composition ing 1988-91 were, however, consis­ November, because only the larger and tently higher than they were during The size ranges of the two halfbeak faster-growing fish are harvested dur­ 1974. Daily catch rates in winter, a peak species overlapped but ballyhoo were ing the summer and early autumn. The fishing season, appeared to be limited typically larger than balao (Fig. 8). In mean size of ballyhoo at the time annu­ principally by vessel capacity, and the this survey, ballyhoo ranged from 108 lus II was formed was 264 mm (1004 slightly larger vessels used in recent mm to 313 mm FL (209 ±26.3; mean ± in) FL (Berkeley and Houde, 1978), a years have allowed for larger winter s.d.; n = 20,307) and balao ranged from size well represented in the fishery dur­ landings. It is even possible that demand 131 mm to 273 mm (194 ± 24.3; n = ing the winter months of 1988-91. Be­ for halfbeaks rose slightly in the late 6,791). After truncating the extreme cause few ballyhoo live longer than two 1980's (Fig. 3), and because no addi­ tails of the length distributions, about years and few balao live longer than one tional boats entered the fishery, each 95% of the ballyhoo harvested were year, the fishery exploited all age classes boat made bigger catches. Berkeley et 150-260 mm (5.9-10.2 in) long and of both species (Berkeley et aI., 1975). aI. (1975) noted that 11 fishing vessels 95% of balao captured were 145-240 operated in the halfbeak fishery in 1974, mm (5.7-9.5 in) long. Ballyhoo also and each vessel carried only four boxes weighed more, on average, than balao for holding fish, whereas during 1988­ (76.8 g vs. 61.2 g), based on year-round 39.7 20 January 91, 10 vessels operated and each usu­ samples from the fishery (n =3,524). 10 ally carried six to eight boxes. By late At least two size classes were evident ,gj~ spring or during the summer, "fresh" for both species (Fig. 8). Small, age-O ~~ ~ :~1~8 ~ F~br~ary bait supplies occasionally become satu­ fish were evident and often dominant I rated and frozen fish are stockpiled, so in July, about four months after spawn­ ~j 0 n= 762 • March boats carry fewer boxes to match the ing begins (Berkeley and Houde, 1978). 10 • n=3041 daily demand. If a haul exceeds the A descriptive growth model, based on >8 0 ~=31~~ April holding capacity of the vessel the fish analysis of scale-determined ages, sug­ ~;jj ~ are released, and this creates small dif­ gests that ballyhoo grows fairly fast :=1321 on=1312~ ferences between actual catch and during its first year and that both spe­ 10 • n=2838 May o>. landed catch values reported. Such oc­ cies reach 209-231 mm (8.2-9.1 ~ ~=~ :j 0 June 20..--r------, ~ ~ .:=165~~i i I >­ Cll_ 'OJ:: :: j 0 n= 785 July ....!.0 ~ ~;.:=1~, Q).­ I 0­ 15 0­ Q)O &. 2O~.~~August >0 10 mOO 38 l:l: 10 .- Cll ~~ ~ ~~~;n1lni... September J::0 j 0::J o il~ :0::0 ..... J:: 5 ~j 0 n= 341~ October J::­ 10 • n=2117 ~ 1;;­u.= I I I I 0 0 n= 0 4.I&:ovember 0 10 j _n=62 _ ,~ 2 3 4 5 6 7 8 9 10 11 12 :J8 - i .," Month ~j ~ ~'~~ i.e,ce~ber

100 125 150 175 200 225 250 275 300 Figure 7.-Comparison of monthly catch rates of halfbeaks (bal­ lyhoo and balao combined) caught during 1988-91 with those Fork length (mm) caught during 1974, as reported in' Berkeley et al. (1975). Data for 1988-91 were converted to numbers of fish per trip based on the average value of 13.7 fish/kg (6.2 fish/lb). There are no standard Figure 8.-Monthly length-frequency of error bars for November because we surveyed only one vessel fish­ ballyhoo and balao, based on pooled data ing during this month in all years. All other months (1988-90) are for 1988-91 and plotted as a stacked his­ represented by >4 fishing-vessel days. Measures of variance were togram. N = number of fish measured. The not reported in Berkeley et al. (1975), and their catch rates for Janu­ ordinate has a maximum of 20% or 30% ary were combined from data collected in 1974 and 1975. for graphical clarity.

36 Marine Fisheries Review Management Implications mented, but Bohnsack et al. (1994) sug­ Information System. B. Collette kindly gested that private anglers in Monroe provided the illustrations for Figure 1. The higher catch rate observed for County harvested 5,000-35,000 S. Berkeley, B. Collette, J. Leiby, M. 1988-91 is the most compelling evi­ halfbeaks (about 400-2,500 kg) annu­ Murphy, J. Quinn, T. Sminkey, and D. dence that the fishery did not overex­ ally during the 1980's. Winkelman offered useful suggestions ploit ha1fbeak populations during the Recent changes in fishing regula­ about the manuscript. 1970's and 1980's. The absence of tions, accomplished through Florida's changes in species composition also constitutional "net ban" referendum Literature Cited supports this conclusion, but this argu­ (enacted July 1995), should affect the Banford, H. M., and B. B. Collette. 1993. ment is less tractable because the fish­ proportions of halfbeak harvested be­ Hyporhamphus meeki, a new species of halfbeak (Teleostei: Hemiramphidae) from the ery does not distinguish between spe­ tween Palm Beach and Monroe coun­ Atlantic and gulf coasts of the United States. cies. The general length-frequency of ties. The surrounding nets used in the Proc. BioI. Soc. Wash. 106(2):369-384. Beets, J., and J. A. LaPlace. 1991. Virgin Islands fish was also similar between 1974 and ballyhoo fishery are now prohibited baitfish survey: October to December 1986. 1988-91, and age-O fish were observed within 1 or 3 miles of shore, on the At­ Proc. Gulf Caribb. Fish. Inst. 40:55-63. in all years examined, suggesting that lantic and Gulf coasts, respectively; Berkeley, S. A., and E. D. Houde. 1978. Biology of two exploited species of halfbeaks, no major year-class failure occurred because the continental shelf adjacent Hemiramphus brasiliensis and H. balao from during this 3-year period. to Palm Beach county is relatively nar­ southeast Florida. Bull. Mar. Sci. 28(4): Estimates of recreational fishing ef­ row, most of the fishing grounds there 624-644. -=-c-' , and F. Williams. 1975. fort indicate that general demand for have been eliminated. Monroe County Fishery and biology of ballyhoo on the south­ bait probably increased little during the (which includes the Florida Keys) has east Florida coast. Univ. Miami Sea Grant Program, Spec. Rep. 4, 15 p. 1980's. In Monroe County specifically, a large number of reefs beyond these Bohlke, J. E., and C. G. Chaplin. 1993. Fishes of the average number of headboats was limits. Monroe County's proportion of the Bahamas and adjacent tropical waters. 2nd relatively constant from 1981 to 1992 the total landings for Florida has already ed. Univ. Tex. Press, Austin, 771 p. Bohnsack, J. A., D. E. Harper, and D. B. (Bohnsack et aI., 1994). The estimated increased from 35% in 1986 to 52% in McClellan. 1994. Fisheries trends from Mon­ total number of recreational fishing trips 1991 and to 75% in 1994, but this oc­ roe County, Florida. Bull. Mar. Sci. made along Florida's Atlantic coast also curred principally because Dade 54(3):982-1018. Broadhead, G. C. 1951. The organization of the did not increase during the 1980's (1980 County's landings have declined pro­ Florida Marine Fisheries Statistical System. = 10.5 million trips; 1985 = 12.5 mil­ portionally. IfPalm Beach landings de­ Proc. Gulf Caribb. Fish. Inst. 3: 114-117. Burgess, G. H., G. W. Link, Jr., and S. W. Ross. lion trips; 1990 = 8.8 million trips; cline as well, then Monroe County's 1979. Additional marine fishes new or rare to Holliday, 1984; NMFS, 1986; Van proportion of landings will increase fur­ Carolina waters. Northeast Gulf Sci. 3(2):74-87. Voorhees et aI., 1992). However, it may ther. Increases in halfbeak landings in Collette, B. B. 1965. Hemiramphidae (Pisces, Synentognathi) from tropical west Africa. be that demand for halfbeaks has grown Monroe County represent both a geo­ Atlantide Rep. 8:217-235. Dan. Sci. Press, while demand for other baitfishes has graphic shift and absolute growth of Ltd., Copenhagen been level or declined. During the late halfbeak landings, because total Florida _---,--,_. 1974. The garfishes (Hemiramphidae) ofAustralia and New Zealand. Rec. Aust. Mus. 1980's some fish processing plants landings have also increased from 29(2): 11-105. switched to flash freezing and vacuum 380,300 kg (836,600 Ib) in 1986 to _=-:--:. 1978. Hemiramphidae.ln W. Fischer packing halfbeaks, and this has helped 523,000 kg (1,150,700 Ib) in 1994. (Editor), FAO species identification sheets for fishery purposes, Western Central Atlantic maintain the quality of halfbeak bait More location-specific information (Fishing Area 31). Vol. II. Food Agric. Organ., (which is particularly important for about the fishery is being collected, and U.N., Rome (unpagin.). -:-:7'--: ' G. E. McGowen, N. V. Parin, and S. trolling the whole fish). If so, then this this will assist us in understanding the Mito. 1984. Beloniformes: development and would increase consumer demand, and current and potential changes in the relationships. In H. G. Moser (Editor in chief), it may explain the apparent increase in fishery. Ontogeny and systematics of fishes, p. 335­ 354. Am. Soc. Ichthyol. Herpetol., Spec. halfbeak landings between 1986 and Acknowledgments Publ. I. 1994 (1991 peak = 0.626 million kg). de Sylva, D. P. 1974. A review of the world sport Continued monitoring of the com­ fishery for billfishes (Istiophoridae and We are grateful to R. DiGiacomo, J. Xiphiidae). In Richard S. Shomura and mercial halfbeak fishery through DiGiacomo, C. Jones, G. Van Munster, Francis Williams (Editors), Proceedings of the Florida's MFIS program and occasional R. Van Munster, Hudgins Fish Co., International Billfish Symposium Kailva-Kona, Hawaii, 9-12 August 1972. Part 2, Review and special surveys are cost-effective for Plantation Fish Co., and all others who contributed papers, p. 12-33. U.S. Dep. tracking halfbeak populations. This ap­ participated in this survey. S. Commer., NOAA Tech. Rep. NMFS-SSRF-675. proach, however, does not monitor Branstetter, F. McIntyre, D. Pierce, and Gentle, E. C. 1977. The charterboat sport fishery of Dade County, Florida, March 1976 to Febru­ halfbeaks harvested by anglers, who use F. Sutter assisted in data collection and ary 1977. Univ. Miami, Coral Gables, 162 p. cast nets, and occasionally small hooks, processing. B. Pollard (National Marine Hardy, J. D., Jr. 1978. Development of fishes of to catch fish that are attracted with Fisheries Service) helped calculate the the mid-Atlantic bight, An atlas of egg, larval and juvenile stages. Vol. 2, Anguillidae oatmeal or chum. The numbers of bal­ effects of inflation on fish prices. We through Syngnathidae. U.S. Dep. Inter., Fish lyhoo harvested directly for bait by an­ also appreciate the efforts of T. Maurer Wild. Serv., BioI. Servo Program, Rep. FWS/ OB5-78/12: 127-138. glers (including those on charter and and J. O'Hop who provided statistical _-;:-;---,-;:;and R. K. Johnson. 1974. Descriptions party boats) has not been well docu­ data from Florida's Marine Fisheries of halfbeak larvae and juveniles from Chesa­

58(1-2}, 1996 37 peake Bay (Pisces: Hemiramphidae). Chesa­ the northeastern Gulf of Mexico during 1971. Stark, W A., Jr., and W. P. Davis. 1966. Night peake Sci. 15:241-246. In Richard S. Shomura and Francis Williams habits of fishes of Alligator Reef, Florida. Hildebrand, S. E, and W C. Schroeder. 1928. (Editors), Proceedings of the International Ichthyol. Aquarium J. 38(4):313-356. Fishes of Chesapeake Bay. Bull. U.S. Bur. Billfish Symposium Kailva-Kona, Hawaii, 9­ Tabb, D. C., and R. B. Manning. 1961. A check­ Fish. XLIII(Pt. I): I52-154. 12 August 1972. Part 2, Review and contrib­ list of the flora and fauna of northern Florida Holliday, M. C. 1984. Marine recreational fish­ uted papers, p. 269-289. U.S. Dep. Commer., Bay and adjacent brackish waters of the ery statistics survey, Atlantic and Gulf coasts, NOAA Tech. Rep. NMFS-SSRF-675.U.S. Florida mainland collected during the period 1979(revised)-1980. US. Dep. Commer., Dep. Commer., NOAA Tech. Rep. NMFS­ July, 1957 through September, 1960. Bull. NOAA, Natl. Mar. Fish. Serv., Curro Fish. Stat. SSRF-675. Mar. Sci. 11(4):552-649. 8322. NMFS. 1986. Marine recreational fishery statis­ Van Voorhees, D. A., J. E Witzig, M. E Osborn, Jolley, J. W, Jr. 1977. The biology and fishery of tics survey, Atlantic and Gulf coasts, 1985. M. C. Holliday, and R. J. Essig. 1992. Marine Atlantic sailfish, Istiophorus platypterus, from U.S. Dep. Commer., NOAA, Natl. Mar. Fish. recreational fishery statistics survey, Atlantic southeast Florida. Fla. Mar. Res. Publ. 28, Serv., Curro Fish. Stat. 8327. and Gulf coasts, 1990-1991. U.S. Dep. 31 p. Nedelec, C. 1982. Definition and classification Commer., NOAA, Natl. Mar. Fish. Serv., Curr. Kimmel, J. J. 1991. A fisheries dependant [sic] offishing gear categories. Food Agric. Organ., Fish. Stat. 9204. survey ofbaitfishes in Puerto Rico. Proc. Gulf U.N., Rome, FAO Fish. Tech. Pap. 222, 51 p. Caribb. Fish. Inst. 40:64-81. Siebenaler, J. B. 1955. Commercial fishing gear Nakamura, E. L., and L. R. Rivas. 1974. An and fishing methods in Florida. Fla. Board analysis of the sportfishery for billfishes in Conserv., Mar. Lab. Tech. Ser. 13,45 p.

38 Marine Fisheries Review