396 



Fishery Bulletin 98(2) 



females from best pots (81.3 ±7.5 mm, n=ll) or worst 

 pots (86.9 ±2.7 mm, n=24) (MWU, Z=1.48, P>0.1). 

 Numbers of male crabs were significantly associated 

 with best pots, whereas female crabs were associ- 

 ated with worst pots ;t'"=53.7, P<0.001); however, 

 this result was highly dependent on two pots with 

 the highest number of males (117) and females (8) 

 respectively. Exclusion of those two pots reduced the 

 j^2 value to insignificance (3.4, P>0.05). Males com- 

 posed 93% of crabs in best pots and 44% in worst 

 pots. This result may be due primarily to the fact 

 that pots in good condition retained large males that 

 could have escaped more easily from damaged pots. 



Even though we could not determine whether holes 

 in webbing were caused by the degradation of bio- 

 degradable twine, recovery, or some other process, 

 we examined their influence on crab CPP. The four 

 highest CPPs for Tanner crabs (10, 17, 22, and 125) 

 occurred in pots without holes, but there was no sig- 

 nificant difference in CPP between pots with intact 

 or torn webbing ( MWU= 1. 17, P>0.26 ). Therefore, we 

 could not reject the hypothesis that pot holes have no 

 effect on CPP. The 95% confidence intervals around 

 the proportion of intact pots containing crabs ( 10 of 

 41; P=0.244, 95% CI range: 0.124-0.403) overlapped 

 those for holed pots with crabs (14 of 106; P=0.132, 

 95% CI range: 0.074-0.211) and included the mean 

 of the latter. Thus the binomial test also failed to 

 reject the null hypothesis. Lack of significance in 

 these tests may be partly due to the high number of 

 pots without crabs in both groups. 



In addition to Tanner crabs, four male red king 

 crabs (Paralithodes camtschaticus) were found in 

 two outside pots. One pot contained 3 legal-size red 

 king crabs with a mean CW of 16 1 mm and a recently 

 dead Pacific cod which probably acted as bait; the 

 other pot contained a single red king crab of 54 mm 

 CW. 



Discussion 



Sidescan sonar has been used to assess the cover- 

 age of fishing grounds by bottom trawlers (Krost et 

 al., 1989) but, to our knowledge, has not previously 

 been used to determine the prevalence of ghost fish- 

 ing by pots or other lost gear. Sonar is a very effective 

 tool in this regard; lost crab pots could easily be iden- 

 tified by their shape and attached lines were often 

 visible on the sonar plots. Occasionally other uniden- 

 tified objects were observed; some of these could have 

 been extremely degraded pots and one looked like 

 the chassis of a truck or shipping van. Since this 

 study was conducted, we have also tested a laser 

 line-scanning system; its resolution (±1 cm) is much 



better than that afforded by sonar and allowed us to 

 identify crabs and other organisms on the bottom as 

 well as inside crab traps (Tracey et al., 1998). 



The mean CPP ( 1.54 crab/pot) for all Tanner crabs 

 does not seem particularly excessive. However, much 

 of these data were obtained from an area where 

 heavy fishing pressure had not occurred for 2.5 years. 

 CPP in the best (i.e. most recently lost) pots was 

 almost seven times that in the worst (i.e. oldest) 

 pots; even though this difference was not statisti- 

 cally significant, it represents an important trend. 

 Thus, examining lost pots soon after an active fish- 

 ery has closed would probably yield high estimates 

 of trapped crabs, and more males of larger sizes. 



In addition to catch rates, it is necessary to know 

 the number of actively fishing ghost pots. One "ball- 

 park" estimate of pot loss rates in the eastern Bering 

 Sea is 20,000 pots per year (Alaska Board of Fisher- 

 ies, cited in Paul et al., 1994), but this estimate was 

 made at a time when up to 100,000 pots were being 

 used in Bering Sea crab fisheries annually. In 1997, 

 the Alaska Board of Fisheries imposed pot limits 

 in the Bering Sea and required all pots to be regis- 

 tered for each crab fishery, partly to reduce loss of 

 pots when weather or ice prevented their recovery. 

 Since then, fewer pots have been used in the Bering 

 Sea, and pot loss rates are presumably lower. In 

 1999, 50,720 pots were registered for the Bering Sea 

 snow crab (C opilio) fishery; preliminary informa- 

 tion indicates that about 1% of these may be lost 

 and replaced during the fishery, as boats return to 

 port for unloading.'^ However, not all lost pots are 

 replaced during intensive fisheries, and during short 

 (<1 week) king or Tanner crab fishery openings, 

 boats do not return to port (a 2-d round trip) or 

 replace pots until after the fishery closes; such losses 

 are not reported. Pot losses as low as 10% per year 

 would contribute 5000 lost pots each year. Subsis- 

 tence pots, on the other hand, are not accounted for, 

 are not required to be registered, and have practi- 

 cally no restricted locations. 



Stevens ( 1996) concluded that the number of active 

 ghost pots would reach a maximum over time, due 

 to the arithmetic increase in lost pots and exponen- 

 tial decay processes. If 7000 pots with a half-Ufe of 4 

 years were lost annually, the number of active ghost 

 pots would stabilize at 44,000 after 40 yr, although 

 most would begin to accumulate by year 25 (Stevens, 

 1996). This is a reasonable time frame for accumula- 

 tion, because pot fishing for crabs in the Bering Sea 

 has been conducted since at least 1966. However, 



Morrison. R. 1999. ADF&G, P.O. Box 308, Dutch Harbor, AK 

 99692. Personal commun. 



