Stevens et al : Ghost fishing by Tanner crab pots 



397 



extrapolation of ghost pot densities in Chiniak Bay 

 to a 40,000 km- area of the Bering Sea (where fish- 

 ing for king. Tanner, and snow crab is particularly 

 intensive) would yield 1.68 million pots in various 

 stages of degradation. The actual number is prob- 

 ably somewhere between these extremes. 



If a reliable estimate of pot loss existed, it would 

 be tempting to speculate on the numbers of crabs 

 killed owing to ghost fishing. However, our estimate 

 of 3.95 Tanner crabs per "best" pot represents only 

 a "snapshot" of the ghost fishing process and does 

 not account for cumulative mortality over time. The 

 difference in CPP between best (3.95) and worst 

 pots (0.58) illustrates that numbers of trapped crabs 

 decline over time, due to escapement, predation, 

 mortality, and reduced capture or retention rates. 

 Starvation may account for a large part of the mor- 

 tality. Kimker (1994) found 39% mortality among 

 legal-size (>139 mm CW) male Tanner crabs that 

 were kept in closed crab pots for 119 days; some of 

 this mortality was undoubtedly due to cannibalism 

 of weaker crabs (Kimker, 1994). Mortality was low 

 for the first 30 days, then increased linearly through 

 the remainder of the experiment. Starvation may be 

 delayed even longer; Paul et al. (1994) held Tanner 

 crabs in laboratory tanks for fixed periods without 

 food, then gave them unlimited access to food. Only 

 10% mortality occurred among crabs during 90 d 

 while they were held without food, but 100% mortal- 

 ity occurred during the following 140 days during 

 which time they had access to food. Overall mortality 

 ranged from 40% for those starved 60 days to 100% 

 for those starved 90 days (Paul et al., 1994), which is 

 consistent with the findings of Kimker (1994). Pre- 

 dation by octopus (High, 1976) and sunflower sea- 

 stars (Breen, 1987) has also been implicated as a 

 cause of mortality of crabs in pots. Octopus were 

 important predators in our study and may cause a 

 significant portion of initial mortality. In a year-long 

 study of blue crab (Callinectes sapidus) ghost fish- 

 ing, Guillory (1993) concluded that, of crabs that 

 were recruited after loss of bait, 51% escaped and 

 45% died. Although mean CPP was only 1.0 (similar 

 to our study ) during weekly pot retrievals, the aver- 

 age total CPP for the entire year was 48 crabs, of 

 which 26 (55%) died (Guillory, 1993). 



Escapement probably accounts generally for the 

 decline in numbers of crabs found in lost pots. 

 Escapement of legal-size (>165 mm CW) king crabs 

 from pots tended toward an asymptotic value of 80%, 

 whereas escapement of smaller crabs was 92% after 

 periods of 14 to 16 d (High and Worlund, 1979). 

 Escapement rate probably depends upon behavioral 

 differences between species. Video observations of 

 crabs in square pots showed that both Tanner and 



king crabs tended to aggregate in the corners of pots, 

 but during random movements some crabs acciden- 

 tally found the entrance tunnel and escaped through 

 it.'^ Tanner crabs are too small to reach the tunnel 

 unless there are large numbers in the pot, allowing 

 them to climb on top of one another to reach the 

 tunnel opening. In addition, degradable twine in 

 Tanner crab pots should create gaps in the webbing 

 within weeks or months after loss, although 34% of 

 our pots had intact webbing even after 1-2 years in 

 the water. That such a large proportion contained 

 no degradable twine suggests that, at a minimum, 

 one third of the commercial crab pots used in the 

 Kodiak region prior to 1994 were illegally con- 

 structed. Apparently, fishermen were taking advan- 

 tage of the very sparse law enforcement presence in 

 Alaskan ports. 



Crabs continue to enter lost pots, even without 

 bait. High and Worlund (1979) showed that king 

 crabs continue to enter unbaited pots for up to 16 

 days. Breen (1987) showed that unbaited traps 

 caught Dungeness crab (Cancer magister) at the 

 same rate one year after becoming lost as they did 

 when first set. He estimated that lost traps caught 

 17 Dungeness crabs per year, of which almost half 

 (9.3) died, and the remainder escaped. Tanner and 

 king crab traps are quite different from Dungeness 

 crab pots, which are made with stainless steel web- 

 bing that degrades very slowly. Nevertheless, Breen's 

 ( 1987 ) study clearly demonstrates that entry, escape- 

 ment, and mortality rates of crabs in lost pots is a 

 dynamic process. 



Seasonality is also an important factor in ghost 

 fishing. Breen (1987) found varying crab numbers 

 in pots at different times of the year and concluded 

 that a study must be conducted all year round to 

 obtain the best estimate of crab ingress rates. Guil- 

 lory ( 1993 ) found that recruitment of blue crabs to 

 ghost pots, mortality, and escapement all varied sea- 

 sonally. High turnover rates can lead to high capture 

 and mortality rates; 2/3 of blue crabs entering traps 

 died or escaped within 2 weeks ( Guillory, 1993 ). Thus 

 it is not possible to estimate mortality from a single 

 observation of recovered traps because the number 

 of crabs in the trap represents a balance of continu- 

 ous ingress, egress, and mortality rates. 



During the 1996 ADF&G annual Gulf of Alaska 

 crab trawl survey, "^ the density of Tanner crabs was 

 approximately 1685 crab/km- in Chiniak Bay, 97 in 

 Kalsin Bay, and 76 in Middle Bay, for an average 



Urban. D. 1997. Bottom trawl survey of crab and ground- 

 fish: Kodiak Island, Chignik, and South Peninsula areas, 

 1996. Regional Information Rep. 4K97-58. Alaska Dept. of Fish 

 and Game, 211 Mission Rd. Kodiak. AK 99615. 



