622 



Fishery Bulletin 103(4) 



Description of gear 



Fleets sampled by observers targeted swordfish (Xiphias 

 gladius), albacore (Thunnus alalunga), or bluefin tuna 

 (Thunnus thynnus). Five fishing gears were studied: 

 swordfish longline (SWO-LL), "American type" swordfish 

 longline (SWO-LL A ), albacore longline (ALB-LL), bluefin 

 tuna longline (BFT-LL), and driftnet (DN). 



The swordfish longline consists of a nylon monofila- 

 ment main line, 2 to 3 mm in diameter, hung in a sag- 

 ging curve between surface floats. Branch lines with 

 a length of 5-18 m descend from the main line, each 

 terminating in a single baited J-hook. The number of 

 hooks ranges from 800 to 2800 and hook size varies 

 from no. to 5. The "American type" swordfish longline, 

 a variation of the aforementioned gear and used mainly 

 in Greece, was introduced in the Greek fishery in the 

 mid 1980s. It consists of fewer hooks (350-700) of size 

 2, much longer branch lines (15-50 m), and a fish at- 

 tractant light stick (Duralumes" Lindgren-Pitman Inc., 

 Pompano Beach, FL) attached to each branch line, 1 m 

 above the bait. The albacore longline is a more lightly 

 constructed longline that has 800 to 4000 J-hooks. hook 

 sizes 6-9, a main line from 1 mm to 1.6 mm in diam- 

 eter, and shorter branch lines (3-6 m). The bluefin 

 tuna longline is the most robust longline, having 1000 

 to 1200 J-hooks of size or 1, a main line 5.0 mm in 

 diameter, and branch lines 45 m long. Frozen mackerels 

 (Scomber scombrus) or (Scomber japonicus) and frozen 

 squids (Illex sp.) or (Loligo sp.) are used as baits, as in 

 the swordfish and bluefin tuna fishery, whereas frozen 

 sardines — Sardina pilchardus or Sardinella sp. — are 

 the baits mainly used in the albacore fishery. Driftnets, 

 ranging from 2.5-20 km in length and from 24-40 m 

 in height and having a stretched mesh size of 380 mm, 

 were used mainly in Italy by the swordfish and tuna 

 fishery. Since 1998, after the enforcement of the regu- 

 latory measures for the driftnets, the traditional nets 

 were rejected and the Italian fishermen introduced a 

 smaller driftnet, called ferrettara. This net has a length 

 of 2.5 km, a depth from 18 to 25 m, and a mesh size 

 of 180 mm. All gears targeting large pelagic fish, both 

 longlines and nets, are shot (deployed) in the evening 

 and their retrieval begins after midnight. Among the 

 gears sampled, the swordfish longline is the main gear 

 used in the Mediterranean Sea. 



Data collection and statistical analyses 



Sampling was carried out during a three-year period 

 from 1998 to 2000. Catch and effort data were derived 

 from records taken by observers stationed both at main 

 fishing ports and onboard 11 commercial fishing vessels, 

 from January 1998 to December 1999. Biological data, 

 such as size and sex of sharks caught, were obtained 

 from January 1998 to September 2000. Observers were 

 present on fishing trips (702 fishing days) and at 17 

 landing sites, performing duties that included collecting 

 fishing and operational data, identifying and measuring 

 fish, as well as recording the exact location and date for 



each fishing set. From each fishing vessel sampled at- 

 sea, these observers collected the following fishing and 

 operational data series: name of fishing boat, gear used, 

 duration of each trip in days, fishing effort per fishing 

 day (number of hooks for longline gear, net length, and 

 depth in meters for driftnet gear), number and weight 

 offish caught per fishing day by species, and number of 

 sharks discarded. Because fishermen generally do not 

 keep reliable logbooks to report their daily catch, sam- 

 pling at landing sites was performed through interviews, 

 as well as by direct observations and measurements. 

 From each boat sampled at the landing site, observ- 

 ers, interviewing fishermen or skippers of the vessels, 

 collected data on the duration of each trip in days, the 

 number of fishing days, the fishing area, and the fishing 

 effort per fishing day. The number and weight of fish 

 landed were observed and measured directly during 

 landing and recorded by species. Biological data for the 

 specimens caught included total length (TL) in cm, fork 

 length (FL) in cm, dressed weight (to the nearest 0.1 kg), 

 and sex when possible. 



To investigate trends in the abundance of sharks, 

 we used the nominal catch per unit of effort (CPUE) 

 expressed as the number of individuals per 1000 hooks 

 for longlines, and per 1000 m of net for driftnets. Fish- 

 ing duration was assumed to be constant because soak 

 time was almost the same for all trips. Setting began at 

 dusk and retrieving began after midnight. Each shark 

 brought onboard vessels was assessed according to the 

 following scale: 1) good — very high motility and ac- 

 tive behavior; 2) fair — moderate motility; 3) poor — poor 

 motility but having the ability to respond to external 

 stimuli; 4) dead or showing no response to external 

 stimuli. 



Chi-square <x 2 ) tests were performed to test varia- 

 tions in species composition by type of fishing gear, 

 area sampled, and by sampling onboard or at landing 

 sites. Catch data were classified in rows (species) and 

 columns (gears, areas sampled, or sampling venue [fish- 

 ing vessel or landing site]) to create contingency tables 

 and were tested for significant association between 

 rows and columns, assuming that row and column clas- 

 sifications are independent (null hypothesis). Nonpara- 

 metric analysis of variance (Kruskall-Wallis test) was 

 performed to compare the total length medians of the 

 samples by fishing gear and per area. Nonparametric 

 analysis of variance was used because our data sets 

 did not meet the criteria needed to use the classical 

 method of analysis of variance (ANOVA) e.g., normally 

 distributed populations, equal variances. 



Results 



A total of 8733 sharks (153.6 t biomass) and 131,912 

 fish of other species (teleosts, rays, and skates) were 

 documented from 5826 fishing days sampled, 5124 at 

 landing sites and 702 onboard, during the two-year 

 period 1998-99 (Tables 1-2). In all areas examined 

 throughout the Mediterranean Sea, sharks represented 



