390 



Fishery Bulletin 98(2) 



vens et al.- High and Worlund (1979) placed red king 

 crab in unbailed square pots; after 16 days, 20% of 

 legal-size crabs remained in the pots. Those crabs that 

 escaped after prolonged enclosure were recaptured at 

 lower rates than those that escaped quickly. Crabs 

 that were placed in closed vinbaited square pots exhib- 

 ited mortahties from 4% to 12% after 16 days. 



In our report we examined ghost fishing by lost 

 crab pots in specific areas off the northeast shore of 

 Kodiak Island, Alaska. In Chiniak Bay, the locations 

 of 189 putative lost pots were discovered in April 

 1994 with sidescan sonar. With this sonar survey as 

 a guide, we estimated the extent of ghost fishing by 

 recovering lost pots and examining their contents. 

 Three studies were conducted: 1) a pilot study within 

 Chiniak Bay in 1995 to develop methods for pot 

 recovery; 2) a targeted study in 1996 to recover spe- 

 cifically identified pots from known locations within 

 Chiniak Bay; and 3 1 a nontargeted study in 1996 

 within Chiniak, Kalsin, Womans, and Ugak Bays, to 

 recover pots from areas known to have been heavily 

 fished during past crab fisheries, without attempt- 

 ing to target specific pots. These efforts were carried 

 out jointly by the Alaska Department of Fish and 

 Game (ADF&G) and the National Marine Fisheries 

 Service (NMFS), with assistance from the U.S. Army 

 Corps of Engineers (USACE) and the Kodiak Island 

 Borough IKIB). 



Materials and methods 



Sidescan sonar survey and in-situ observations 



The sonar survey was conducted by scientists at 

 SAIC, Inc., using a dual frequency (100/500 KHz) 

 sidescan sonar system (Klein model 595), with both 

 a paper chart and digital data signal processor. Posi- 

 tions were recorded with a differential GPS receiver. 

 The sonar was operated from the Alaska Department 

 of Fish and Game (ADF&G) RV Resolution during 

 6-8 April 1994, in a portion of Chiniak Bay at depths 

 of 100-150 m. Transects 3 km in length were run at 

 a spacing of 75 m, such that the scanning of each 

 line overlapped those on each side by 50%, provid- 

 ing a total coverage of 150%. A total area of approxi- 

 mately 4.5 km- was surveyed. From 22 April to 2 

 May 1995, the two-person submersible Delta was 

 used to survey the area for a study of crab behavior 

 (Stevens et al., in press); pots that were encountered 



2 Stevens, B.C.. J. A. Haaga, and W.E.Donaldson. 1993. Under- 

 water observations on behavior of king crabs escaping from crab 

 pots. Alaska Fisheries Science Center processed report 9.3-06. 

 Alaska Fisheries Science Center, NMFS, 7600 Sand Point Way 

 NE, Seattle, WA 98115, 14 p. 



accidentally during that study were examined. On 

 27 April 1995, a remotely operated vehicle (model 

 S2 ROV Phantom, Deep Ocean Engineering, Inc.) 

 equipped with a sector-scanning sonar was also used 

 to locate and examine crab pots. 



1995 pilot study 



Pots identified from the sidescan sonar survey were 

 plotted on a map (Fig. 1), then numbered. A random 

 sample of 23 single pots was chosen for trial recov- 

 ery in April 1995. A short (7.6-m) grappling chain, 

 designed to hook the pot or its floating line, or both, 

 was towed from the main boom of the RV Resolu- 

 tion. The grappling chain was lowered into the water 

 when the ship was approximately 250 m from the 

 estimated position of the targeted pot, then towed 

 past the position at less than 5.6 km/h (3 kni. If 

 the recovery operation was not successful, the vessel 

 would continue to tow the grappling chain in a circle 

 around the same position until the pot was hooked 

 or until we gave up. When a pot was hooked, it was 

 brought onboard with the ship's crane. Information 

 about the pot and its contents was recorded, includ- 

 ing pot type, condition, damage to the frame or web- 

 bing, and presence of bait jars and biodegradable 

 mesh. In most cases the grapple snagged the steel 

 frame of the pot and it was not possible to determine 

 if snagging action caused additional damage to the 

 webbing. 



In our paper, we use the terms "trap" and "pot" 

 interchangeably. The three types of crab traps most 

 commonly used in the Gulf of Alaska are square, 

 pyramidal, and conical ( see High and Worlund, 1979, 

 for detailed descriptions). Square pots are the most 

 commonly used type in the Bering Sea and are made 

 with steel frames, usually >2 x 2 m, with two funnel- 

 shaped entrance tunnels on opposite sides. Pyrami- 

 dal and conical pots are slightly smaller and have a 

 single square or round entrance at the top, which is 

 usually fitted with a plastic collar. All three of these 

 types of pots are covered with nylon twine webbing. 

 Cod pots are square pots that have been modified for 

 cod fishing by the addition of restriction devices to 

 prevent entry of crabs and escapement of cod. Sub- 

 sistence pots, used to capture crabs for personal con- 

 sumption, may be any of the above styles of pots or 

 variations thereof Pot conditions were rated on a 

 subjective five-point scale from "very poor" to "excel- 

 lent." Criteria used to evaluate condition included 

 the condition of webbing (presence of holes or major 

 gaps), numbers of broken frames and corner joints, 

 and estimated age of the pot. Pots coded as "good" 

 or "excellent" were assumed to have been lost less 

 than one year and were probably placed for subsis- 



