Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Haunted waters : an estimate of ghost fishing of crabs and lobsters by traps Poon, Amy Min-Yee 2005

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2005-0299.pdf [ 6.93MB ]
Metadata
JSON: 831-1.0074876.json
JSON-LD: 831-1.0074876-ld.json
RDF/XML (Pretty): 831-1.0074876-rdf.xml
RDF/JSON: 831-1.0074876-rdf.json
Turtle: 831-1.0074876-turtle.txt
N-Triples: 831-1.0074876-rdf-ntriples.txt
Original Record: 831-1.0074876-source.json
Full Text
831-1.0074876-fulltext.txt
Citation
831-1.0074876.ris

Full Text

H A U N T E D W A T E R S : A N E S T I M A T E O F G H O S T F I S H I N G O F C R A B S A N D L O B S T E R S B Y T R A P S b y A M Y M I N - Y E E P O O N B . S c . , The U n i v e r s i t y o f B r i t i s h C o l u m b i a , 1999 A T H E S I S S U B M I T T E D I N P A R T I A L F U L F I L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F M A S T E R O F S C I E N C E i n T H E F A C U L T Y O F G R A D U A T E S T U D I E S (Resource Management and Env i ronmen ta l Studies) T H E U N I V E R S I T Y O F B R I T I S H C O L U M B I A M a r c h 2005 © A m y P o o n , 2005 Abstract: Ghost fishing is the mortality caused by lost fishing gear and as such, its effect on target populations is difficult to quantify. This study seeks to rectify that by developing a methodology to estimate the effect of ghost fishing by lost pots and traps on large crustaceans (crabs and lobsters). Forty-four case studies, consolidated from published literature on the amount of lost traps and their mortality rates, are used as the basis of the calculations. Missing parameters are inferred from existing values in similar case studies. A Monte Carlo simulation is used to analyze the sensitivity of the results to uncertainties in the input data. The mean estimate of ghost fishing as a percentage of reported landings, over the forty-four case studies, is 3 . 8 % . Ghost fishing is a management issue to contend with, especially in areas where fishing effort is high. Improved legislation on using biodegradable twine and/or Galvanic Time Release mechanisms has and will ameliorate the destructive impact of lost traps on vulnerable ecosystems. 11 Table of Contents: Abstract i i Table o f Contents i i i L i s t o f Tables i v L i s t o f F igures v A c k n o w l e d g e m e n t s v i i C H A P T E R I Introduction 1 C H A P T E R II M e t h o d o l o g y 5 2.1 Case studies chosen 5 2.2 F lowchar t o f calculat ions 6 2.3 Inferences on miss ing parameters 8 2.4 M o n t e C a r l o s imulat ions 8 C H A P T E R III Case Study Resul ts 11 3.1 True crabs 13 3.2 K i n g crabs 79 3.3 Lobsters 95 3.4 S p i n y lobsters 114 C H A P T E R I V D i s c u s s i o n and C o n c l u s i o n 118 References 114 A p p e n d i x 1 V i s u a l B a s i c code used i n M o n t e C a r l o s imulat ions 132 i n List of Tables: Table 1: Number of case studies with missing parameters 8 Table 2: Comparison of F-test results 9 Table 3: Comparison of estimates by taxonomic group 121 i v List of Figures: Figure 2.1: Input parameters and f low chart o f calculat ions 7 F igu re 3.1: Phylogenet ic tree o f studied species 12 F igure 3.1.1: B r i t i s h C o l u m b i a Dungeness crab fishery 1984 74 F igure 3.1.2: C a l i f o r n i a Dungeness crab fishery 1986 74 F igure 3.1.3: N e w B r u n s w i c k snow crab fishery 1987 74 F igu re 3.1.4: G u l f o f St. L a w r e n c e snow crab fishery 1966-1989 74 F igure 3.1.5: G u l f o f St. L a w r e n c e snow crab fishery 1987 75 F igure 3.1.6: Wash ing ton State Dungeness crab fishery 1975-1976 75 F igure 3.1.7: L o u i s i a n a blue crab fishery c. 1999 75 F igure 3.1.8: L o u i s i a n a blue crab fishery c. 1993 75 F i g u r e 3.1.9: L o u i s i a n a b lue crab fishery 1988 75 F igure 3.1.10: N e w f o u n d l a n d snow crab fishery 1974 75 F igure 3.1.11: Pac i f i c Coast Dungeness crab fishery 1971-1972 76 F igure 3.1.12: Wash ing ton State Dungeness crab fishery 1972 76 F igure 3.1.13: Wash ing ton State Dungeness crab fishery 1973 76 F igure 3.1.14: A l a b a m a blue crab fishery 1999-2000 76 F igure 3.1.15: Eastern B e r i n g Sea crab fishery 1993 76 F igure 3.1.16: Eastern B e r i n g Sea snow crab fishery 1999 76 F igure 3.1.17: F l o r i d a blue crab fishery 2001 77 F igure 3.1.18: G u l f o f M e x i c o blue crab fishery 1993 77 F igure 3.1.19: G u l f o f M e x i c o blue crab fishery 1999 or 2000 77 F igure 3.1.20: G u l f o f St. L a w r e n c e snow crab fishery 1980 77 F igure 3.1.21: L o u i s i a n a blue crab fishery 1999 or 2000 77 F igure 3.1.22: L o u i s i a n a blue crab fishery 1993 77 F igure 3.1.23: Ireland spider crab fishery 2000 78 F igure 3.1.24 Ireland b r o w n crab fishery 1968-1972 78 F igure 3.1.25: Texas blue crab fishery 1999 or 2000 78 F igure 3.1.26: M i s s i s s i p p i blue crab fishery 1999 or 2000 78 F igure 3.1.27: South C a r o l i n a blue crab fishery c. 1979 78 F igure 3.1.28: Texas blue crab fishery c. 1997 78 F igure 3.2.1: C h i n i a k B a y red k i n g crab fishery c. 1990s 94 F igure 3.2.2: Eastern B e r i n g Sea red k i n g crab fishery c. 1993 94 F igure 3.2.3: B r i s t o l B a y red k i n g crab fishery c. 1994 94 F igure 3.2.4: B r i s t o l B a y red k i n g crab fishery c. 1996 94 F igure 3.2.5: B r i s t o l B a y red k i n g crab fishery 1990 94 F igure 3.2.6: B r i s t o l B a y red k i n g crab fishery 1991 94 F igure 3.2.7: A l a s k a red k i n g crab fishery 1969-1970 95 F igure 3.3.1: A t l an t i c offshore lobster fishery 1976 113 F igure 3.3.2: A t l an t i c seacoast inshore lobster fishery 1973 113 F igure 3.3.3: M a i n e lobster fishery 1960 113 v Figure 3.3.4: N e w E n g l a n d lobster fishery 1987 113 F igure 3.3.5: M a i n e lobster fishery 1975 114 F igure 3.3.6: M a i n e lobster fishery 1992 114 F igure 3.3.7: M a i n e and Massachusetts lobster f ishery 1991 114 F igu re 3.3.8: A t l an t i c lobster fishery 1988 114 F igure 3.4.1: H a w a i i a n spiny lobster fishery 1989 .' 117 F igure 4 .1 : Frequency chart o f ghost f ishing as a percentage reported landings 121 v i Acknowledgements: The author w o u l d l ike to thank her thesis advisor , D r . D a n i e l Pauly , for his unremit t ing guidance and support, as w e l l as the members o f her thesis committee, D r . T o n y Pi tcher , D r . R e g Watson , and D r . L e s L a v k u l i c h for their encouragement and valuable input. D r . Rober t Ahrens was instrumental i n deve lop ing the V i s u a l B a s i c code used i n the M o n t e C a r l o s imulat ions . v i i Chapter I Introduction W h i l e the ocean 's resources were once thought to be inexhaustible, reports that bot tom fisheries were dec l in ing and catches were d i m i n i s h i n g were beg inn ing to be publ i shed i n the early 1900s (Garstang ci ted i n C h o p i n et al. 1997). The f i sh ing industry m o v e d into the 2 0 t h century w i t h the realisation that the abundance o f f ish was l imi ted and that f i sh ing practises must be moderated for the industry to continue. Y e t , most f ish stocks today are s t i l l fu l ly exploi ted or severely depleted. M i l l i o n s o f dol lars are poured into the science that manages these stocks, but un t i l recently, on ly a piece o f the puzz le has been studied i n detai l . Fisheries scientists tend to focus on f ishing morta l i ty f rom commerc i a l landings and, to a lesser degree, art isinal and recreational landings. A s a result, many techniques employed b y fisheries scientists to prevent overf i sh ing, such as Ind iv idua l Transferable Quotas or l imi ted f i sh ing access t ime, focus on l i m i t i n g the landings; that is , the k n o w n morta l i ty f rom f i sh ing gear. H o w e v e r , landings compr i se o n l y one por t ion o f overa l l f i sh ing mortal i ty . C h o p i n et al. (1997) list the other causes o f mor ta l i ty due to f ishing as: i l l ega l and misreported landings; mor ta l i ty associated w i t h discards; mor ta l i ty associated w i t h f ish after escape f rom f ishing gear; mortal i ty associated w i t h ghost f ishing; mor ta l i ty associated w i t h f ish pass ively dropping o f f or out o f f i sh ing gears; morta l i ty associated w i t h f ish avo id ing the f ishing gear; morta l i ty associated w i t h predat ion after escape; arid morta l i ty due to changes i n habitat associated w i t h f ishing. Because many o f these f ishing- induced mortal i t ies are indirect, it is diff icul t to assess their impact on f ish populat ions. H o w e v e r , as studies have shown, they are significant enough to warrant some examinat ion . F o r example, A l v e r s o n et al. (1997) estimated that the average amount o f byca tch is 3 0 % o f total reported landings, a l though an update b y K e l l e h e r ( in press) lowers it to 8%. W h e n most f ish resources are fished at or over m a x i m u m capaci ty ( C h o p i n et al. 1997), this is s t i l l a significant amount. T h i s paper focuses o n the impacts o f mor ta l i ty associated w i t h ghost f ishing, spec i f ica l ly that o f crabs and lobsters. W h i l e the phrase 'ghost f i sh ing ' might b r i n g to m i n d the haunting v i s i o n o f d rowned fishers gu id ing incorporeal nets i n the depths o f the ocean, the reali ty is not so poetic; it is , however , just as c h i l l i n g . Storms, accidents, or l i t tering at sea cause lost gear that may continue to capture fish for some t ime. F o r example , 42,000 to 63,000 lobster pots are lost every year i n N e w Eng land , corresponding to 20 -30% o f the gear that is used ( S m o l o w i t z 1978b). W h e n lobsters are caught i n these traps, they die from starvation or predation i f they cannot escape. The i r decompos ing bodies act as bait for other lobsters, w h i c h are d rawn to the trap and are captured as w e l l . T h i s v ic ious cyc le continues for the effective life o f the trap, w h i c h can last up to 15 years ( H i g h and W o r l u n d i n Stevens 1996). 1 F o r the purposes o f this paper, I w i l l be us ing the f o l l o w i n g terms and defini t ions: • L o s t gear refers to f ishing gear that is lost by fishers, either through theft or at sea. • G h o s t gear refers to f ishing gear that "continue f ishing after a l l cont ro l o f that gear is lost by the f isherman" - ( S m o l o w i t z 1978b) • D e r e l i c t gear is gear that can no longer fish due to damage, (after B r e e n 1990) Ghos t f i sh ing , then, is the r emova l o f f ish from a popula t ion by lost f ishing gear, but the def in i t ion o f ghost gear appears to be f lex ib le i n literature. Some authors use the terms ' los t gear' and 'ghost gear' interchangeably, and others use one to mean the other. H o w e v e r , not a l l lost gear is capable o f ghost f i sh ing; for instance, gear can be lost due to thefts or degradation. T h o u g h concern has been raised over the impacts o f lost gear, few studies have been done to quantify them. Cer ta in ly , there have been few attempts to consolidate the i nd iv idua l studies to see the overa l l picture, and no attempts to examine ghost f i sh ing on a g loba l scale. H o w e v e r , ghost f i sh ing is an issue that is present i n any fishery that uses passive gears - especia l ly when the species most l i k e l y to be affected are those o r ig ina l ly targeted by the ghost gear ( B r o w n 1998). F i s h i n g gear can be lost i n a variety o f ways . B u o y l ines that mark the locat ions o f traps can be cut b y vessel traffic, frayed through negligence or age, tangled i n storms, towed away b y gillnetters or trawlers, cut del iberately i n disputes among fishers, lost i f their locat ions are not proper ly recorded, or s i m p l y misplaced . In N o r w a y ( G o d e y 2003) and N e w B r u n s w i c k , Canada (Ma l l e t et al. 1988), drif t ing ice causes considerable loss o f traps. A d d i t i o n a l l y , traps m a y be abandoned, especia l ly by fishers w h o are o n l y i n the fishery for the short-term. In a survey o f b lue crab fishers i n the G u l f o f M e x i c o , 2 5 % o f the ind iv idua l s in te rv iewed fished s ix months or less ( G u i l l o r y et al. 2001) . U p l i n e s m a y break dur ing haul ing . Lobster fishers w h o do not ac t ively f ish dur ing the win te r sometimes 'wet store' coated-wire traps at sea, result ing i n greater losses f rom winter storms and boat traffic (Carr and Har r i s 1987). A l l these lost traps m a y ghost fish for the rest o f their effective l ives , and fishers are charged nothing for this damage, nor does the loss fo rm part o f their quota. T h e pe r iod i n w h i c h ghost gear r emain active is affected b y the mater ia l f rom w h i c h the trap is made, its locat ion, and the season i n w h i c h it is lost. In N o r t h A m e r i c a , legis la t ion requires many crab fisheries to use G a l v a n i c T i m e d Releases ( G T R s ) or biodegradable plast ic on traps to render them inoperable after a set t ime, shortening their ghost f i sh ing per iod i f lost. Traps that do not use G T R s continue to fish for up to 15 years 2 ( H i g h and W o r l u n d 1979), un t i l destroyed by storm damage, bur ia l i n sand, corros ion, woodborers , or other vagaries. I f traps are dragged b y boat traffic, m o v e d b y storms o f f the continental shelf, or ' m o v e d into areas where they are u n l i k e l y to attract prey, their ghost f ishing effectiveness w o u l d be l o w . L i k e w i s e , traps that are lost off-season w o u l d catch litt le prey — but i f they were s t i l l intact by the next season, their ghost f ishing potential w o u l d be fu l f i l led . Once gear is lost, it can interact w i t h prey through a variety o f mechanisms. Ghos t traps m a i n l y capture the o r ig ina l target species, but m a y catch others as w e l l . F o r example , ghost tanner crab traps i n K o d i a k , A l a s k a , captured not o n l y tanner crabs, but also red k i n g crabs, pacif ic octopus, rockf i sh , arrowtooth flounder, sunflower stars, and sea cucumbers (Stevens et al. 2000) . A n i m a l s that have been trapped i n ghost gear usua l ly starve to death, turn to cann iba l i sm o f n e w l y mol ted crustaceans, or make easy prey for outside predators (Breen 1990). The i r dead bodies often serve as bait for addi t ional captures, as w i t h lobsters. E v e n i f the animals manage to escape the traps, they m a y have morta l wounds or weakened defences that lower their chances for su rv iva l after their escape. Regardless o f the mechan i sm, ghost f ishing mor ta l i ty reduces avai lable popula t ion biomass, and thus impacts o n reproduct ion, and therefore o n f ishing. In l ight o f the potential damage caused by ghost gear, var ious measures have been taken to min ima l i s e its effects. L e g i s l a t i o n has been put into place to require the use o f p roper ly marked buoys, escape gaps, and degradable plast ics or ga lvanic t imed releases. In A l a s k a , a l l traps used at sea are required to have a panel that w i l l open after the destruction o f degradable fibre (Breen 1985); as are traps i n Puerto R i c o and the V i r g i n Islands (Paul 1984). Federal l aw i n the U n i t e d States mandates escape vents and a degradable panel that w o u l d a l l o w lega l s ized lobsters and most other prey to escape ghost pots (Carr and Har r i s 1987). Some fisheries use 'habipots ' that lobsters can use as shelter and enter and leave unhindered ( S m o l o w i t z 1978b). Pre-emptive prevention o f ghost f i sh ing by decreasing the number o f traps lost is another method for reducing the damage caused by lost gear. A s s i g n i n g trap-free lanes i n areas o f heavy boat traffic prevents buoy lines from be ing tangled or cut. The most effective w a y o f reducing trap loss w o u l d be to reduce f i shing effort. E v e n g iven the m y r i a d methods used to decrease the impact o f ghost f ishing, it remains an issue i n the f ishing industry. R e d u c i n g f ishing effort is diff icul t and expensive to enforce (Breen 1990), and escape ports ameliorate but do not el iminate the p r o b l e m o f ghost f ishing (Stevens et al. 1993). W h i l e some fisheries i n N o r t h A m e r i c a are l ega l ly mandated to fit escape gaps or biodegradable panels on traps, this is not necessar i ly true for the rest o f the w o r l d . F o r example, there is no such legis la t ion i n the 3 U n i t e d K i n g d o m other than by l aws i n two areas: the C u m b r i a Sea Fisheries and States o f Jersey ( B u l l i m o r e 2001). Furthermore, even fisheries that do use biodegradable panels and galvanic t imed releases cannot be sure o f the p rec i s ion o f the devices ( K i m k e r 1994) and the traps are capable o f act ively ghost f ishing for the per iod before they are made ineffective. Cer ta in mou lded plast ic escape panels i n lobster traps m a y be large enough to release lobsters, but not larger fish that enter the trap (Carr and Har r i s 1987). M e a n w h i l e , gear manufacturers are developing more durable traps that w o u l d last longer for fishers, but also have longer ghost f i sh ing l ives i f lost. A recent advertisement boasts o f a maintenance-free trap that "does not rust, w o n ' t rot, is imperv ious to marine borers, and is capable o f wi ths tanding the r igours o f many h i g h l y product ive seasons." ( T N F O F I S H 1995) G i v e n that ghost f i sh ing does have an impact o n fisheries resources despite the preventive actions taken to reduce their damage, it behooves us to determine just h o w m u c h damage is done. B y consol ida t ing and standardising informat ion f rom publ i shed studies on ghost f i sh ing and propos ing a methodology to estimate the amount o f ghost f i sh ing i n an area, this thesis strives to quantify ghost f i sh ing act iv i ty relative to reported landings. M i s s i n g values were inferred f rom s imi la r , established values to complete the data set. M o n t e C a r l o s imulat ions were then run to p rov ide confidence l imi ts for the results. For ty-four case studies were examined i n total and presented wi th a parameter list , a narrative to exp la in the source o f the values used, and a summary o f variances used i n the M o n t e C a r l o s imulat ions . The thesis concludes w i t h a d iscuss ion o f uncertainties i n the methodology and poss ible ways to alleviate the impact o f ghost f ishing. 4 Chapter II Methodology The p rob lem w i t h assessing the impact o f indirect f i sh ing mortal i ty such as ghost f i sh ing is that it is , b y nature, diff icul t to direct ly observe. H o w does one locate 5,000 lost cod pots i n N e w f o u n d l a n d and record a l l the organisms they k i l l o n an annual basis? M o r t a l i t y caused b y ghost f i sh ing is too expensive to assess on a large scale, so they are studied i n experiments, w i t h the assumption that the results obtained from contro l led experimentat ion are representative o f what occurs i n the ocean. T o examine the extent o f damage caused b y lost gear, scientists need to k n o w h o w m u c h gear is lost and h o w m u c h morta l i ty each gear causes. Surveys f rom fishers can be used to get an idea o f the number o f gear that is lost i n a trip and the estimate can be used as a factor for the results f rom the experiments. C o u n t i n g the number o f lost pots observed i n a f ixed area us ing underwater vehic les is an alternate method ( C h o p i n et al. 1997). Capture rates and species mortal i ty are studied b y s imula t ing lost gear i n a contro l led area, where they are regular ly checked for v ic t ims . Crab pots i n A l a s k a have been studied most extensively, where researchers check catches by br ing ing the pots up onto boats and setting them again or b y sending scuba divers to determine mor ta l i ty in situ. Studies on ghost f i sh ing have been done i n var ious parts o f the w o r l d , spec i f ica l ly i n N o r t h A m e r i c a , but there have been few reviews o f those studies. B reen (1990) and C h o p i n et al. (1997) summarised a number o f studies that estimated the numbers o f lost gear. Th i s thesis consolidates and standardises informat ion found i n various studies o f ghost f ishing i n reg iona l waters i n order to make estimates o f ghost f ishing mor ta l i ty and compare different fisheries i n quantitative terms. Ghos t f i sh ing mortal i ty i n var ious case studies is presented as a percentage o f the reported catch, and the methodology is out l ined be low. 2.1 C a s e s tudies chosen Fisher ies used i n the case studies were chosen o n the basis o f avai lable information. A literature search was done to look for studies on recover ing lost gear or f ie ld experiments on the mor ta l i ty o f animals i n ghost traps, as these two parameters are specific to ghost f i sh ing. The informat ion f rom the studies was then consol idated to fo rm case studies o f 44 fisheries, 36 o f w h i c h were i n 13 U S states ( inc lud ing one i n H a w a i i ) , 6 i n Canada, and 2 in Ireland. Case studies that were mi s s ing parameters c ruc ia l to ident i fy ing them as unique fisheries, such as total gear or total reported landings, or specific to the target species, such as species weight , were not used. 5 2.2 F l o w c h a r t o f c a l cu l a t i ons In order to obtain an estimate o f ghost f i sh ing i n a fishery relative to its reported landings, the f o l l o w i n g parameters were extracted f rom publ ished articles: • N u m b e r o f gear active i n a fishery (by number o f fishers, boats, or gear) • Rate o f loss per fishery or per boat (as a percentage o f total number o f gear) • A m o u n t o f gear lost per year (either p u l l e d f rom the literature or calculated f rom the above two parameters) • Percentage o f lost gear assumed to be ghost f i sh ing (as a percentage o f lost gear) • T i m e dur ing w h i c h a pot is ac t ively ghost f i sh ing (number o f years) • M o r t a l i t y rate o f animals i n lost pots (average number o f animals per pot) • A v e r a g e weight o f species caught i n the pots (ki lograms) • Repor ted landings o f the species i n quest ion i n the part icular f ishery i n that year (tonnes); i f no year was specif ied, then the year o f the publ ica t ion was used. The numbers were used to compute annual amounts and rates, without account ing for seasonal var iab i l i ty , o w i n g to pauci ty o f studies cons ider ing this factor. T h e number o f traps i n an area and the animals caught are assumed to be at equ i l ib r ium. Recrea t iona l fisheries were not inc luded , due to pauci ty o f data and di f f icul ty i n determining h o w often recreational fishers worked , al though the f i shing effort can be considerable i n some cases. F igure 2.1 demonstrates h o w the parameters are used to determine the amount o f ghost f ishing. Parameters correspond to the number ing system used i n V i s u a l B a s i c code for the uncertainty analysis, w h i c h is expla ined i n Sect ion 2.3. ' N u m b e r o f f ishers ' and ' N u m b e r o f traps per f isher ' are label led differently because they are not used i n the code: their product ( 'To ta l gear') is used rather than the ind iv idua l parameters. T h i s a l lows the code to beg in i n a standard pos i t ion whether total gear is g iven i n a reference or it has to be calculated from number o f fishers and number o f traps per fisher. 6 Number of fishers (Parameter i) X Number of traps / fisher (Parameter ii) Total gear (Parameter 1) Annual trap loss rate (Parameter 2) Gear lost per year (Parameter 3) Percentage of lost traps ghost fishing (Parameter 4) Number of traps ghost fishing (Parameter 5) Length of ghost fishing period (Parameter 6) Number of traps ghost fishing per year (Parameter 7) Mortality rate per trap (Parameter 8) Annual ghost fishing mortality (Parameter 9) Average weight of species (Parameter 10) Weight of annual ghost fishing mortality (Parameter 11) Reported landings in fishery (Parameter 12) Figure 2.1: Flowchart of calculations (calculated parameters in octagons) Percentage of ghost fishing relative to reported landings (Result; PerGhostCalc in VB code) 7 2.3 Inferences on missing parameters Some publ i shed articles p rov ided one or more parameters estimated; others concentrated on a specific parameter. I f a l l the required parameters were not avai lable i n a g iven case study, the mi s s ing parameter was obtained from an average o f s imi la r fisheries for w h i c h estimates were avai lable . I f no s imi la r fishery existed, the grand mean o f a l l case studies w i t h a value for that parameter was taken. E a c h case study was then ranked i n order o f re l i ab i l i ty o f its data, its rank depending on h o w many parameters it suppl ied. O f the 44 case studies, 1 study p rov ided a l l eight key parameters (see above); 5 were mi s s ing 1; 17 were mi s s ing 2; 16 were mi s s ing 3, and 5 were m i s s i n g 4 (Table 1). N o t e that al though some case studies were mi s s ing the same number o f parameters, it was not necessari ly the same parameters that were mi s s ing for a l l o f them. Table 1: Number of case studies with missing parameters Number of missing parameters (out of 8) Number of case studies applicable 0 1 1 5 2 17 3 16 4 5 Severa l assumptions had to be made w h e n implement ing this approach, notably that i f certain parameters m a y vary i n different studies, a l l other factors are the same. A l s o , f ie ld experiments done over a short per iod at a specific loca t ion were assumed to be representative o f annual condi t ions i n a larger region, and the number o f animals i n a ghost trap is assumed to be at equ i l i b r i um w h e n counted. S ince not a l l studies specif ied whether a l l o f the lost gear ended up ghost f i shing, I took the average o f the values i n the case studies where the percentage o f lost gear actual ly ghost f i sh ing was g iven , unless specif ied otherwise. 2.4 Monte Carlo Simulations D u e to uncertainties i n the parameters, a M o n t e C a r l o analysis was done on each case study to obtain confidence l imi t s for the ghost f ishing estimates. M i s s i n g parameters inferred from values i n other case studies ( in section 2.3) were g iven distributions based on those from the other studies us ing the formula: 8 S d ( ( X + Y ) / n ) = ( sq r t (Var (X) + V a r ( Y ) ) ) / n where n is the number o f values used to infer the mis s ing parameter. In the M o n t e C a r l o analyses, values were g iven as confidence intervals, w h i c h were expressed as percentages o f their absolute values. W h e n uncertainty was expressed i n the form o f standard errors, these were doubled to generate approximate confidence intervals. W h e n a range is presented rather than a mid-poin t estimate, the mid-range was used, w i t h the upper and lower l imi t s o f the range expressed as a percentage o f the median. Where no measure o f var iab i l i ty was avai lable i n a g iven study, the mean uncertainty for the parameter i n question i n s imi la r fisheries was appl ied, or the grand mean was used i f estimates were unavai lable for s imi la r fisheries. W i t h upper l imi t s , midranges, and lower l imi t s avai lable to describe the confidence interval o f each parameter, a symmetr ica l tr iangular distributions was assumed. T h i s dis t r ibut ion gives equal s ignif icance to the three values used i n the distr ibutions and el iminates the p r o b l e m posed b y the asymptot ic ta i l i n a normal dis tr ibut ion. M o n t e C a r l o s imulat ions were run for each o f the case studies to get a dis t r ibut ion o f possible ghost f i sh ing estimates. A V i s u a l B a s i c program was wri t ten for this process us ing data f rom a M i c r o s o f t E x c e l spreadsheet ( A p p e n d i x 1). M o n t e C a r l o s imulat ions were sampled 10,000 times each. In order to determine h o w many samples were required for the M o n t e Ca r lo s imulat ions , a two-sample F-test o n va ry ing number o f runs was conducted on M o n t e C a r l o results f rom the South C a r o l i n a blue crab fishery w h i c h was among the f ive case studies w i t h four mi s s ing parameters and therefore, presumably , the most uncertainty. The trials were run twice and their variances compared. T h e probabi l i ty o f F be ing smaller than the cr i t ica l va lue for var ious numbers o f runs is as fo l lows : Table 2: Comparison of F-test results N u m b e r o f r u n s P r o b a b i l i t y o f F < = c r i t i c a l v a l u e 10 0.051 100 0.456 1000 0.339 10000 0.390 The n u l l hypothesis assumes that the variances o f the two populat ions (or tr ial runs, i n this case) are equal; on ly w h e n F is less than or equal to the F-test statistic's cumulat ive dis t r ibut ion function ( C D F ) is the n u l l hypothesis rejected. O n l y the trials i n w h i c h the M o n t e C a r l o s imulat ions were sampled 10 times had a 9 5 % probabi l i ty o f rejecting the 9 n u l l hypothesis. E v e n sampl ing 100 times is sufficient i n that the probabi l i ty that the two tr ial runs d id not differ s ignif icant ly enough i n their results to reject the n u l l hypothesis. A d d i t i o n a l l y , E f r o n and T ibsh i r an i (1986) state that 1,000 samples is an approximate m i n i m u m to compute reasonably accurate confidence intervals. Sampl ing 10,000 times a l lows the graphing o f a smoother frequency dis t r ibut ion. 10 CHAPTER III Case Study Results Resul ts f rom the methodology appl ied to the case studies are presented be low. N o t e the repeti t ion o f procedure i n the case studies; this a l lows each case study to stand i n d i v i d u a l l y wi thout the need to allude to others for references. Case studies are ordered by taxonomic group and then b y number o f mi s s ing parameters. Species groupings were chosen f o l l o w i n g the assumption that different species act differently i n their capture behaviour , w h i c h w o u l d impact their morta l i ty rate i n lost traps. F igure 3.1 shows the phylogenet ic relat ionships between the species studied i n this thesis (based o n informat ion f rom R a w l i n g s et al. 2004). 11 Decapoda Brachyura (True crabs) Anomura Astacidea Palinura (King crabs) ~• (Lobsters) ~"—" (Spiny lobsters)1 Cancroidea Majoidea Portunoidea Paguroidea Nephropoid Palinuroidei Figure 3.1: Phylogenetic tree of species studied Cancridae Majidae Cancer | Portunidae —• Lithodidae — Nephropidae' Palinuridae — Chionoecetes Maia ——— Callinectes -Paralithodes " Ho mar us r :;ri".'.:v.; —— Cancer magister Cancer pagurus Chionoecetes opilio Maja squinado Callinectes sapidus Paralithodes camtschatica Homarus amehcanus Panulrius marginatus 12 3.1 T r u e c r abs C a s e s tudies w i t h 0 m i s s i n g p a r a m e t e r s : British Columbia Dungeness crab fishery 1984 Parameter list The f o l l o w i n g parameters were used i n the calculat ions demonstrated i n the f low chart i n F igure 2.1. Italics denote that the value was calculated f rom previous parameters. Parameter i : 104 vessels X Parameter i i 252.4 ± 2 6 % traps per vessel = Parameter 1 26,250 traps X Parameter 2 10.9% = Parameter 3 2,861 traps X Parameter 4 5 3 . 5 % = Parameter 5 1,531 traps X Parameter 6 2.2 years = Parameter 7 3,338 traps X Parameter 8 9.3 +7%/-15% crabs- trap"'year = Parameter 9 31,319 crabs X Parameter 10: 0.68 k g = Parameter 11: 21,297 kg -I- Parameter 12: 292 t = Resul t : 7.3% Narrative A l l data for the Dungeness crab {Cancer magister) f ishery i n the Fraser R i v e r Estuary, B r i t i s h C o l u m b i a , were obtained f rom Breen (1987). In 1983, 104 vessels i n the fishery used an average o f 252 traps per vessel , so the total number o f traps i n use was estimated to be 26,250. The trap loss rate was estimated to be 10.9%, result ing in 2,861 traps lost annually. 53 .5% o f those traps were actually ghost f i sh ing , and their lifespan was estimated at 2.2 years, so the number o f traps ghost f ishing, assuming that the system was at equ i l ib r ium, w o u l d be 3,338. E a c h trap caused a mor ta l i ty o f 9.3 crabs per year; the total annual mortal i ty is thus 31,319 crabs. W i t h a mean weight per crab o f 0.68 k g , the ghost f ishing catch becomes 21,297 k g , 7 .3% o f the 292 t o f the reported landed i n 1984. 13 Monte Carlo simulation Breen (1987) estimated that each vessel used an average o f 252.4 traps where the 9 5 % confidence l imi t is 186-319, or 252.4 ± 26%. The mor ta l i ty rate was 9.3 crabs- trap' 'year" 1 where the 9 5 % confidence l imi t is 7.88 -9.95 crabs- trap"'year"', o r 9.3 +7%/-15% crabs- t rap" 'year 1 . The dis tr ibut ion o f ghost f i sh ing mortal i ty as a percentage o f the reported landed catch is shown i n F igure 3.1.1. Case studies with 1 missing parameter: California Dungeness crab fishery 1986 Parameter list The f o l l o w i n g parameters were used i n the calculat ions demonstrated i n the f l o w chart i n F igure 2 .1 . Italics denote that the value was calcula ted f rom previous parameters. Parameter i : Parameter i i : Parameter 1: Parameter 2: Parameter 3: Parameter 4: Parameter 5: Parameter 6: Parameter 7: Parameter 8: Parameter 9: Parameter 10: Parameter 11: Parameter 12: Resul t : N o t p rov ided N o t p rov ided N o t p rov ided N o t p rov ided 100,000 traps lost 9 6 . 9 % + 3 .1%/- 51 .3% 96,853 traps 0.375 ± 1 1 % years 36,320 traps 12.9 crabs- trap"1 year"' 466,970 crabs 0.68 k g 317,539 kg 3,520,873 k g 9.0% Narrative The Ca l i fo rn ian Department o f F i s h and G a m e estimates a loss o f 100,000 Dungeness crab (Cancer magister) traps per year ( K e n n e d y 1986). There was no ind ica t ion o f whether a l l the traps were ghost f ishing. T o achieve a conservative estimate, 14 it was assumed that not a l l the lost traps w o u l d ghost f ish, and a modif ier o f 9 6 . 9 % was appl ied. Th i s modifer was the mean calculated from other case studies where this parameter was specif ied: 4 6 . 5 % i n the 1987 B C Dungeness crab fishery (Breen 1987); 100% i n the 1990 C h i n i a k B a y red k i n g crab fishery (assumed from R . M e y e r , pers c o m m to S m o l o w i t z 1978a); 100% i n the 1976 At l an t i c offshore lobster fishery (assumed from S m o l o w i t z 1978b); 100% i n the 1969-1989 G u l f o f St. Lawrence snow crab fishery (assumed f rom V i e n n e a u and M o r i y a s u 1994); 100% i n the 1987 G u l f o f St. L a w r e n c e snow crab fishery ( M a l l e t et al. 1998); 100% i n the 1987 N e w B r u n s w i c k snow crab fishery (Ma l l e t et al. 1998); 100% i n the L o u i s i a n a b lue crab fishery c. 1993 ( G u i l l o r y 1999); 100% i n the 1988 L o u i s i a n a blue crab fishery ( G u i l l o r y 1999); 100% i n the 1993 L o u i s i a n a blue crab fishery ( G u i l l o r y 1999); and 100% i n the 1974 N e w f o u n d l a n d snow crab fishery ( M i l l e r 1977). The result was 96,853 ghost fish traps. T o ameliorate potential losses o f crab populat ions f rom ghost f i shing, the C a l i f o r n i a Dungeness crab fishers are us ing cotton w e b b i n g that renders lost traps inoperable after approximate ly 4-5 months. T a k i n g the midrange f rom this estimate and assuming that the rate o f entry o f lost traps was equal to the rate o f degradation, 36,320 traps are estimated to be ghost f ishing inshore i n the area annually. In 1970, Dah l s t rom (1975 ci ted i n Pac i f i c M a r i n e Fisheries C o m m i s s i o n 1978) retr ieved 90 crabs i n 7 crab traps i n Ca l i f o rn i a ; the average is 12.9 crabs per trap. The length o f t ime i n w h i c h these traps had been ghost f i sh ing is u n k n o w n ; for the purposes o f this case study, it is assumed that they had been f ishing for one year and that the number o f caught crabs was i n equ i l ib r ium - that is , the number o f crabs taken f rom the avai lable popula t ion at any g iven t ime. W h e n the mor ta l i ty rate is appl ied to the number o f ghost f ishing traps, the annual morta l i ty is 466,970 crabs. T o express this mor ta l i ty as a percentage o f the reported landings, the average weight for Dungeness crab (0.68 kg ; (Breen 1987) was appl ied. The weight o f Dungeness crab morta l i ty f rom ghost traps was therefore estimated to be 317,539 k g , or 9 .0% o f the reported landings i n C a l i f o r n i a i n 1986 (3,520,873 k g ; N a t i o n a l M a r i n e Fisher ies Service 2004). Monte Carlo Simulation The percentage o f traps lost to ghost f i sh ing is the grand mean calculated f rom case studies where the parameters were g iven , w h i c h results i n 96 .9% + 3 .1%/- 5 1 . 3 % o f lost traps (96 .9% ± 51 .3%, w i t h the upper l imi t truncated at 100%). K e n n e d y (1986) estimated the l ifespan o f a ghost f ishing trap at 4-5 months, w h i c h translates to 0.375 years ± 11%. The dis t r ibut ion o f ghost f ishing mor ta l i ty as a percentage o f the reported landed catch is shown i n F igure 3.1.2. 15 New Brunswick snow crab fishery 1987 Parameter list Parameter i : 81 vessels (not used i n calculat ion) Parameter i i : N o t p rov ided Parameter 1: N o t p rov ided Parameter 2: N o t p rov ided Parameter 3: 2,466 traps X Parameter 4: 100% = Parameter 5: 2,466 traps X Parameter 6: 0.417 + 141%/-100% years = Parameter 7: 1,029 traps X Parameter 8: 44.3 crabs- trap"'year" 1 = Parameter 9: 45,581 crabs X Parameter 10: 0.576 k g = Parameter 11: 26,238 kg Parameter 12: 7,341,092 k g = Resul t : 0.36% Narrative In 1987, 81 vessels i n the N e w B r u n s w i c k snow crab (Chionoecetes opilio) fishery he ld l icenses and lost between them a total o f 2,466 traps dur ing the season ( M a l l e t et al. 1998). In their o w n calculat ions, M a l l e t et al. (1998) assumed that a l l traps w o u l d be capable o f ghost f ishing; because this case study is der ived f rom that publ ica t ion , the same assumption was made here. H o w e v e r , no estimate for the functional l i fespan o f the traps was avai lable , so an average length o f t ime, 0.417 years, was taken f rom a l l case studies i n w h i c h ghost f i sh ing per iod was specified. The values f rom those case studies were: 2.2 years i n the 1984 B C Dungeness crab fishery (Breen 1987); 0.418 years i n the 1969-70 A l a s k a red k i n g crab fishery ( K i m k e r 1990); 0.375 years ± 1 1 % i n the 1986 Ca l i fo rn i a Dungeness crab fishery (Kennedy 1986); 0.245 +44%/-19% years i n the 1992 C o o k Inlet red k i n g crab fishery ( K i m k e r 1990); and 0.243 ± 3 3 % years. The last value is a composi te i n itself, obtained f rom 0.241 +20.8%/-27.1% years ( K i m k e r 1990) and 0.245 +44%/-19% years ( K i m k e r 1990), and used i n 9 case studies: 1993 B e r i n g Sea crab fishery; 1993 B e r i n g Sea red k i n g crab fishery; 1990 B r i s t o l B a y red k i n g crab fishery; 1991 B r i s t o l B a y red k i n g crab fishery; 1994 B r i s t o l B a y red k i n g crab fishery; 1996 B r i s t o l B a y red k i n g crab fishery; C h i n i a k B a y red k i n g crab fishery i n the 1990s; and 2 1991 eastern B e r i n g Sea snow crab fisheries. 16 A s s u m i n g that the rate o f entry o f lost traps was equal to the rate o f degradation, 1,029 traps are estimated to be ghost fishing i n N e w B r u n s w i c k annually. In a t rawl survey o f the G u l f o f St. Lawrence , M a l l e t et al. (1998) found a mean o f 44.3 crabs per lost trap, w h i c h takes the annual ghost fishing morta l i ty to 45,581 ind iv idua l s . M a l l e t et al. (1998) equate 44.3 ind iv idua ls per trap w i t h 25.5 k g per trap, w h i c h suggests that the average weight o f snow crabs i n that system is 0.576 k g . Therefore, the morta l i ty from ghost traps is 26,238 k g . S n o w crab landings i n the southwestern G u l f o f St. Lawrence amounted to 11,782,000 k g i n 1987 ( M a l l e t et al. 1998) and the N e w B r u n s w i c k fleet had 81 o f the 130 l icensed fishers i n the southwestern G u l f o f St. Lawrence . A s s u m i n g the landings are proportionate throughout the fishery, N e w B r u n s w i c k ' s catch is 7,341,092 k g . The propor t ion o f ghost f i sh ing catch w i t h relat ion to reported landings is thus 0.36 % . M a l l e t et al. (1998), without taking into account the functional l i fe o f a lost trap, estimated the catch lost to ghost f i sh ing to be 62,900,000 k g , or 0 .36% o f N e w B r u n s w i c k ' s reported catch. Monte Carlo Simulation The values for the l i fespan o f ghost f i sh ing traps i n this case study was an average o f thirteen other case studies, some o f w h i c h had asymmetr ica l variances. In the cases where variances were asymmetr ica l , the average o f the variances i n each case study was taken and used towards the f o l l o w i n g formula : S d ( ( X + Y ) / 1 3 ) = ( sqr t (Var (X) + V a r ( Y ) ) ) / 1 3 resul t ing i n a mean o f 0.417 + 141%/-100% years (0.417 ± 141%, w i t h the lower l imi t truncated at 0). The dis t r ibut ion o f ghost fishing mor ta l i ty as a percentage o f the reported landed catch is shown i n F igure 3.1.3. Gulf of St. Lawrence snow crab fishery 1966-1989 Parameter list Parameter i : N o t p rov ided Parameter i i : N o t p rov ided Parameter 1: N o t p rov ided Parameter 2: N o t p rov ided Parameter 3: 831 traps X Parameter 4: 100% = Parameter 5: 831 traps X Parameter 6: 0 . 417+ 141%/-100% years 17 X X Parameter 7: Parameter 8: Parameter 9: Parameter 10 Parameter 11: Parameter 12 Resul t : 347 traps 41 crabs- trap"'year 14,216 crabs 0.572 k g 8,132 kg 11,782,000 0.069% Narrative Between 1966 and 1989, 19,110 traps were lost i n the G u l f o f St. Lawrence snow crab (Chionoecetes opilio) f ishery (V ienneau and M o r i y a s u 1994). A s s u m i n g that the rate o f loss was un i fo rm throughout the 23 years, 831 traps w o u l d have been lost annual ly . The estimate o f trap loss i n v o l v e d on ly traps lost at sea, so it is assumed that they were a l l capable o f ghost f i sh ing. N o estimate for the functional l ifespan o f the traps was avai lable , so an average length o f t ime, 0.417 years, was taken from a l l case studies i n w h i c h ghost f i sh ing per iod was specif ied. The values from those case studies were: 2.2 years i n the 1984 B C Dungeness crab fishery (Breen 1987); 0.418 years i n the 1969-70 A l a s k a red k i n g crab fishery ( K i m k e r 1990); 0.375 years ± 1 1 % i n the 1986 C a l i f o r n i a Dungeness crab fishery (Kennedy 1986); 0.245 +44%/-19% years i n the 1992 C o o k Inlet red k i n g crab fishery ( K i m k e r 1990); and 0.243 ± 3 3 % years. T h e last value is a composi te i n itself, obtained from 0.241 +20.8%/-27.1% years ( K i m k e r 1990) and 0.245 +44%/-19% years ( K i m k e r 1990), and used i n 9 case studies: 1993 B e r i n g Sea crab fishery; 1993 B e r i n g Sea red k i n g crab fishery; 1990 B r i s t o l B a y red k i n g crab fishery; 1991 B r i s t o l B a y red k i n g crab fishery; 1994 B r i s t o l B a y red k i n g crab fishery; 1996 B r i s t o l B a y red k i n g crab f ishery; C h i n i a k B a y red k i n g crab fishery i n the 1990s; and 2 1991 eastern B e r i n g Sea snow crab fisheries. A s s u m i n g that the rate o f entry o f lost traps was equal to the rate o f degradation, 347 traps are estimated to be ghost f i sh ing i n the G u l f o f St. L a w r e n c e annually. V i e n n e a u and M o r i y a s u (1994) determined that the mor ta l i ty rate o f crab caught b y a lost trap is 5 0 % that o f regular ly funct ioning traps; appl ied to the observed C P U E i n the Spr ing 1993 fishery o f 82 ind iv idua l s per trap, the 347 lost traps w o u l d k i l l an average o f 14,216 snow crabs each year. V i e n n e a u and M o r i y a s u (1994), based on unpubl ished data f rom Y . Chiasson , D F O G u l f R e g i o n , state that the average i nd iv idua l weight is 0.572 k g . T h e weight o f ghost f i sh ing catch is thus 8,132 k g , 0 .069% o f the reported 11,782,000 k g landed i n 1987 (Ma l l e t et al. 1998). Monte Carlo Simulation The values for the l i fespan o f ghost f ishing traps i n this case study was an average o f thirteen other case studies, some o f w h i c h had asymmetr ica l variances. In the cases 18 where variances were asymmetr ica l , the average o f the variances i n each case study was taken and used towards the f o l l o w i n g formula: S d ( ( X + Y ) / 1 3 ) = ( sq r t (Var (X) + Var (Y) ) ) /13 resul t ing i n a mean o f 0.417 + 141%/-100% years (0.417 ± 141%, w i t h the lower l imi t truncated at 0). The dis t r ibut ion o f ghost f ishing mor ta l i ty as a percentage o f the reported landed catch is shown i n F igure 3.1.4. Gulf of St. Lawrence snow crab fishery 1987 Parameter list Parameter i : 130 vessels (not used i n calcula t ion) Parameter i i : N o t p rov ided Parameter 1: N o t p rov ided Parameter 2: N o t p rov ided Parameter 3: 3,958 traps X Parameter 4: 100% ' -= Parameter 5: 3,958 traps X Parameter 6: 0.417 +141 % / - 1 0 0 % years = Parameter 7: 1,651 traps X Parameter 8: 44.3 crabs- trap^'year"1 = Parameter 9: 45,581 crabs X Parameter 10: 0.576 k g = Parameter 11: 42,110 kg -7- Parameter 12: 11,782,000 k g = Result : 0.36% Narrative The N e w B r u n s w i c k snow crab (Chionoecetes opilio) fishery, w h i c h invo lves eighty-one o f the 130 l icensed fishers i n the southwestern G u l f o f St. Lawrence ( inc lud ing N e w B r u n s w i c k , Quebec, and Cape B r e t o n Island), lost 2,466 traps i n 1987 (Mal l e t et al. 1998). A s s u m i n g a proport ional rate o f loss i n the southwestern G u l f o f St. Lawrence fishery, 3,958 traps w o u l d have been lost i n that area. In their o w n calculat ions for the amount o f ghost f i sh ing i n the N e w B r u n s w i c k snow crab fishery, M a l l e t et al. (1998) assumed that a l l traps w o u l d be capable o f ghost f i sh ing; because this case study is der ived from that publ ica t ion , the same assumption was made here. H o w e v e r , no 19 estimate for the functional l ifespan o f the traps was avai lable , so an average length o f t ime, 0.417 years, was taken f rom a l l case studies i n w h i c h ghost f ishing per iod was specified. The values from those case studies were: 2.2 years in the 1984 B C Dungeness crab fishery (Breen 1987); 0.418 years i n the 1969-70 A l a s k a red k i n g crab fishery ( K i m k e r 1990); 0.375 years ± 1 1 % i n the 1986 C a l i f o r n i a Dungeness crab fishery (Kennedy 1986); 0.245 +44%/-19% years i n the 1992 C o o k Inlet red k i n g crab fishery ( K i m k e r 1990); and 0.243 ± 3 3 % years. The last va lue is a composi te i n itself, obtained from 0.241 +20.8%/-27.1% years ( K i m k e r 1990) and 0.245 +44%/-19% years ( K i m k e r 1990), and used i n 9 case studies: 1993 B e r i n g Sea crab fishery; 1993 B e r i n g Sea red k i n g crab fishery; 1990 B r i s t o l B a y red k i n g crab fishery; 1991 B r i s t o l B a y red k i n g crab fishery; 1994 B r i s t o l B a y red k i n g crab fishery; 1996 B r i s t o l B a y red k i n g crab fishery; C h i n i a k B a y red k i n g crab fishery i n the 1990s; and 2 1991 eastern B e r i n g Sea snow crab fisheries. A s s u m i n g that the rate o f entry o f lost traps was equal to the rate o f degradation, 1,651 traps are estimated to be ghost f i sh ing i n the G u l f o f St. Lawrence annual ly . In a t rawl survey o f the G u l f o f St. Lawrence , ( M a l l e t et al. 1998) found a mean o f 44.3 crabs per lost trap, w h i c h takes the annual ghost f i sh ing mor ta l i ty to 73,155 ind iv idua l s . M a l l e t et al. (1998) equate 44.3 ind iv idua l s per trap w i t h 25.5 k g per trap, w h i c h suggests that the average weight o f snow crabs i n that system is 0.576 k g . Therefore, the mor ta l i ty f rom ghost traps is 42,110 k g , or 0 .36% o f the reported 11,782,000 k g landed i n 1987 (Ma l l e t et al. 1998). M a l l e t et al. (1998) estimate the morta l i ty from ghost traps at 125,800 k g , w h i c h w o u l d make it 1.7% o f the reported catch. Monte Carlo Simulation The values for the lifespan o f ghost f i sh ing traps i n this case study was an average o f thirteen other case studies, some o f w h i c h had asymmetr ica l variances. In the cases where variances were asymmetr ica l , the average o f the variances i n each case study was taken and used towards the f o l l o w i n g formula: S d ( ( X + Y ) / 1 3 ) - ( sqr t (Var (X) + V a r ( Y ) ) ) / 1 3 result ing i n a mean o f 0.417 + 141%/-100% years (0.417 ± 141%, w i t h the lower l imi t truncated at 0). The dis t r ibut ion o f ghost f ishing mor ta l i ty as a percentage o f the reported landed catch is shown i n F igure 3.1.5. 20 C a s e s tudies w i t h 2 m i s s i n g p a r a m e t e r s : Washington State Dungeness crab fishery 1975-1976 Parameter list Parameter i : N o t p rov ided Parameter i i : N o t p rov ided Parameter 1: N o t p rov ided Parameter 2: N o t p rov ided Parameter 3: 6,577 traps X Parameter 4: 96 .9% + 3 .1%/- 5 1 . 3 % o f lost traps ghost f ishing = Parameter 5: 6,370 traps ghost fishing X Parameter 6: 0.417 + 141%/-100% years Parameter 7: 2,658 traps / Parameter 8: 114 crabs- t rap" 'yeaf 1 = Parameter 9: 303,162 crabs X Parameter 10: 0.68 k g = Parameter 11: 206,150 kg -7- Parameter 12: 3,716,753 k g = Resul t : 5.6% Narrative In the 1975/1976 season, Dungeness crab (Cancer magister) fishers i n Wash ing ton lost an estimated 6,577 traps (Nor thup 1978 ci ted i n M u i r et al. 1984). There was no ind ica t ion o f whether a l l the traps were ghost f ishing. T o achieve a conservative estimate, it was assumed that not a l l the lost traps w o u l d ghost f ish , and a modi f ie r o f 96 .9% was applied. Th i s modifer was the mean calculated f rom other case studies where this parameter was specif ied: 4 6 . 5 % i n the 1987 B C Dungeness crab fishery (Breen 1987); 100% in the 1990 C h i n i a k B a y red k i n g crab fishery (assumed from R . M e y e r , pers c o m m to S m o l o w i t z 1978a); 100% i n the 1976 At l an t i c offshore lobster fishery (assumed f rom S m o l o w i t z 1978b); 100% i n the 1969-1989 G u l f o f St. Lawrence snow crab fishery (assumed from V i e n n e a u and M o r i y a s u 1994); 100% i n the 1987 G u l f o f St. L a w r e n c e snow crab fishery ( M a l l e t et al. 1998); 100% i n the 1987 N e w B r u n s w i c k snow crab fishery (Mal l e t et al. 1998); 100% in the L o u i s i a n a blue crab fishery c. 1993 ( G u i l l o r y 1999); 100% i n the 1988 L o u i s i a n a blue crab fishery ( G u i l l o r y 1999); 100% i n the 1993 L o u i s i a n a b lue crab f ishery ( G u i l l o r y 1999); and 1 0 0 % i n the 1974 N e w f o u n d l a n d snow crab fishery ( M i l l e r 1977). The result was 6,370 ghost fish traps. 21 N o estimate for the functional l ifespan o f the traps was avai lable , so an average length o f t ime, 0.417 years, was taken f rom a l l case studies in w h i c h ghost f i sh ing per iod was specif ied. The values from those case studies were: 2.2 years i n the 1984 B C Dungeness crab fishery (Breen 1987); 0.418 years i n the 1969-70 A l a s k a red k i n g crab fishery ( K i m k e r 1990); 0.375 years ± 1 1 % i n the 1986 Ca l i fo rn i a Dungeness crab fishery ( K e n n e d y 1986); 0.245 +44%/-19% years i n the 1992 C o o k Inlet red k i n g crab fishery . ( K i m k e r 1990); and 0.243 ± 3 3 % years. The last value is a composi te i n itself, obtained f rom 0.241 +20.8%/-27.1% years ( K i m k e r 1990) and 0.245 +44%/-19% years ( K i m k e r 1990), and used i n 9 case studies: 1993 B e r i n g Sea crab fishery; 1993 B e r i n g Sea red k i n g crab fishery; 1990 B r i s t o l B a y red k i n g crab fishery; 1991 B r i s t o l B a y red k i n g crab fishery; 1994 B r i s t o l B a y red k i n g crab fishery; 1996 B r i s t o l B a y red k i n g crab fishery; C h i n i a k B a y red k i n g crab fishery i n the 1990s; and 2 1991 eastern B e r i n g Sea snow crab fisheries. A s s u m i n g that the rate o f entry o f lost traps was equal to the rate o f degradation, 2,658 traps are estimated to be ghost f i sh ing inshore i n the area annual ly . In a 1982 study i n the C o l u m b i a R i v e r estuary, M u i r et al. (1984) used four standard c o m m e r c i a l crab traps to examine the escape o f Dungeness crabs i n s imulated lost traps. O v e r a 28-day per iod , 35 crabs d ied i n the four traps (an average o f 8.75 crabs per trap). A s s u m i n g the same morta l i ty rate throughout the year, the year ly morta l i ty can be extrapolated to 114 ind iv idua l s per trap, result ing i n 303,162 ind iv idua l s caught i n a l l lost traps annually. T o express this morta l i ty as a percentage o f the reported landings, the average weight for Dungeness crab (0.68 k g ; B r e e n 1987) was appl ied. The weight o f Dungeness crab mor ta l i ty f rom ghost traps was therefore estimated to be 206,150 k g , or 5.6% o f the averaged landings for Wash ing ton State averaged between 1975 and 1976 (3,716,753 k g ; N a t i o n a l M a r i n e Fisheries Service 2004) . Monte Carlo Simulation The percentage o f traps lost to ghost f i sh ing is the grand mean calculated f rom case studies where the parameters were g iven , w h i c h results i n 9 6 . 9 % + 3 .1%/- 51 .3% o f lost traps (96.9% ± 51 .3%, w i t h the upper l imi t truncated at 100%). The values for the l ifespan o f ghost f i sh ing traps i n this case study was an average o f thirteen other case studies, some o f w h i c h had asymmetr ica l variances. In the cases where variances were asymmetr ica l , the average o f the variances i n each case study was taken and used towards the f o l l o w i n g formula : S d ( ( X + Y ) / 1 3 ) = ( sq r t fVar (X) + V a r ( Y ) ) ) / 1 3 22 resul t ing i n a mean o f 0.417 + 141%/-100% years (0.417 ± 141%, w i t h the lower l imi t truncated at 0). The dis tr ibut ion o f ghost f ishing morta l i ty as a percentage o f the reported landed catch is shown i n F igure 3.1.6. Louisiana blue crab fishery c. 1999 Parameter list Parameter i : Parameter i i : Parameter 1: Parameter 2: Parameter 3: Parameter 4: Parameter 5: Parameter 6: Parameter 7: Parameter 8: Parameter 9: Parameter 10: Parameter 11: Parameter 12: Resul t : N o t p rov ided N o t p rov ided 500,000 traps 1 0 % 50,000 traps 1 0 0 % 50,000 traps 0.417 + 141%/-100% years 20,862 traps ghost fishing annually 53.8 ± 3 4 . 1 % crabs- trap"'year" ! 1,122,388 crabs 0.16 k g 178,190 kg 18,897,136 k g 0.94% Narrative In Lou i s i ana , approximate ly 500,000 c o m m e r c i a l b lue crab {Callinectes sapidus) traps are i n use annual ly ( G u i l l o r y 1999). I f 10% o f those traps are lost, approximate ly 50,000 traps are added i n L o u i s i a n a each year ( G u i l l o r y 1999). G u i l l o r y (1999) assumes that a l l these lost traps are ghost f i shing, but gives no estimate for the durat ion o f their functional l ives , so the average (0.417 years) was taken f rom other case studies where this parameter was specif ied and appl ied. The values f rom those case studies were: 2.2 years i n the 1984 B C Dungeness crab fishery (Breen 1987); 0.418 years i n the 1969-70 A l a s k a red k i n g crab fishery ( K i m k e r 1990); 0.375 years ± 1 1 % i n the 1986 C a l i f o r n i a Dungeness crab fishery ( K e n n e d y 1986); 0.245 +44%/-19% years i n the 1992 C o o k Inlet red k i n g crab fishery ( K i m k e r 1990); and 0.243 ± 3 3 % years. The last value is a composi te i n itself, obtained f rom 0.241 +20.8%/-27.1% years ( K i m k e r 1990) and 0.245 +44%o/-19% years ( K i m k e r 1990), and used i n 9 case studies: 1993 B e r i n g Sea crab 23 fishery; 1993 B e r i n g Sea red k i n g crab fishery; 1990 B r i s t o l B a y red k i n g crab fishery; 1991 B r i s t o l B a y red k i n g crab fishery; 1994 B r i s t o l B a y red k i n g crab fishery; 1996 B r i s t o l B a y red k i n g crab fishery; C h i n i a k B a y red k i n g crab fishery i n the 1990s; and 2 1991 eastern B e r i n g Sea snow crab fisheries. A s s u m i n g that the rate o f entry o f lost traps was equal to the rate o f degradation, 20,862 traps are estimated to be ghost f ishing i n the area annually. In G u i l l o r y ' s (1993) study o f ghost f ishing by blue crab traps i n Lou i s i ana , an average o f 25.8 ind iv idua l s died per trap. Arcemen t and G u i l l o r y (1993) performed a compar i son o f morta l i ty o f crabs i n traps that had escape r ings and those that had no escape rings. In the former, an average o f 5.3 ind iv idua l s were k i l l e d per trap i n a 2-month per iod. In the latter, an average o f 17.3 ind iv idua l s were k i l l e d per trap i n a 2-month per iod. A s s u m i n g that morta l i ty rate is un i fo rm throughout the study per iod and throughout the year, annual mor ta l i ty is and 31.8 and 103.8 crabs per trap respect ively. The average o f these three values is 53.8 crabs- trap^year" 1 . A p p l i e d to the previous parameter, the annual mor ta l i ty f rom ghost traps is 1,122,388 crabs. The average weight o f b lue crabs i n October 1999 to N o v e m b e r 2000 landings was 0.35 lbs, or 0.16 k g ( M u r p h y et al. 2001). T h e weight o f ghost f i sh ing catch is thus 178,190 k g , or 0 .94% o f the reported 18,897,136 k g landed i n 2001 (Nat iona l M a r i n e Fisheries Service 2004). Monte Carlo Simulation The values for the l i fespan o f ghost f ishing traps i n this case study was an average o f thirteen other case studies, some o f w h i c h had asymmetr ica l variances. In the cases where variances were asymmetr ica l , the average o f the variances i n each case study was taken and used towards the f o l l o w i n g formula: S d ( ( X + Y ) / 1 3 ) = ( sqr t fVar (X) + V a r ( Y ) ) ) / 1 3 result ing i n a mean o f 0.417 + 141%/-100% years (0.417 ± 141%, w i t h the lower l imi t truncated at 0). The morta l i ty rate is taken as the mean from three values, 53.8 crabs-trap"'year" 1 ± 3 4 . 1 % , as expla ined i n the narrative above. The dis t r ibut ion o f ghost f ishing morta l i ty as a percentage o f the reported landed catch i s shown i n F igure 3.1.7. 24 Louisiana blue crab fishery c. 1993 Parameter list Parameter i : N o t p rov ided Parameter i i : N o t p rov ided Parameter 1: 650,000 traps X Parameter 2: 1 0 % = Parameter 3: 65,000 traps X Parameter 4: 100% = Parameter 5: 65,000 traps X Parameter 6: 0.417 + 141%/-100% years: = Parameter 7: 27,121 traps ghost fishing annually X Parameter 8: 53.8 ± 3 4 . 1 % crabs- trap _ 1year"' = Parameter 9: 1,459,104 crabs X Parameter 10: 0.16 k g = Parameter 11: 231,647 kg Parameter 12: 18,897,136 k g = Resul t : 1.2% Narrative In the L o u i s i a n a b lue crab (Callinectes sapidus) f ishery, over 650,000 traps are used ( G u i l l o r y et al. 199'4 ci ted i n Arcement and G u i l l o r y 1994). G u i l l o r y (1999) estimates an annual trap loss rate o f 10% i n the fishery; 65,000 traps w o u l d be added to the ghost f ishing fleet each year. G u i l l o r y (1999) assumes that a l l these lost traps are ghost f ishing, but gives no estimate for the durat ion o f their functional l ives , so the average (0.417 years) was taken from other case studies where this parameter was specified. The values f rom those case studies were: 2.2 years i n the 1984 B C Dungeness crab fishery (Breen 1987); 0.418 years i n the 1969-70 A l a s k a red k i n g crab f ishery ( K i m k e r 1990); 0.375 years ± 1 1 % i n the 1986 C a l i f o r n i a Dungeness crab fishery (Kennedy 1986); 0.245 +44%/-19% years i n the 1992 C o o k Inlet red k i n g crab f ishery ( K i m k e r 1990); and 0.243 ± 3 3 % years. The last va lue is a composi te i n itself, obtained from 0.241 +20.8%/-27.1% years ( K i m k e r 1990) and 0.245 +44%/-19% years ( K i m k e r 1990), and used i n 9 case studies: 1993 B e r i n g Sea crab fishery; 1993 B e r i n g Sea red k i n g crab fishery; 1990 B r i s t o l B a y red k i n g crab fishery; 1991 B r i s t o l B a y red k i n g crab fishery; 1994 B r i s t o l B a y red k i n g crab fishery; 1996 B r i s t o l B a y red k i n g crab fishery; C h i n i a k B a y red k i n g crab fishery i n the 1990s; and 2 1991 eastern B e r i n g Sea snow crab fisheries. A s s u m i n g that the rate o f entry o f lost traps was equal to the rate o f degradation, 27,121 traps are estimated to be ghost f i sh ing i n the area annual ly . 25 In G u i l l o r y ' s (1993) study o f ghost f i sh ing by blue crab traps i n L o u i s i a n a , an average o f 25.8 ind iv idua ls died per trap. Arcemen t and G u i l l o r y (1993) performed a compar i son o f morta l i ty o f crabs i n traps that had escape rings and those that had no escape rings. In the former, an average o f 5.3 ind iv idua ls were k i l l e d per trap i n a 2-month per iod . In the latter, an average o f 17.3 ind iv idua l s were k i l l e d per trap i n a 2-month per iod . A s s u m i n g that morta l i ty rate is un i fo rm throughout the study pe r iod and throughout the year, annual morta l i ty is and 31.8 and 103.8 crabs per trap respect ively. The average o f these three values is 53.8 crabs- t r a p " ' y e a r '• A p p l i e d to the previous parameter, the annual mortal i ty f rom ghost traps is 1,459,104 crabs. The average weight o f b lue crabs i n October 1999 to N o v e m b e r 2000 landings was 0.35 lbs, or 0.16 k g ( M u r p h y et al. 2001). The weight o f ghost f i sh ing catch is thus 231,647 k g , or 1.2% o f the reported 18,897,136 k g landed i n 2001 (Na t iona l M a r i n e Fisher ies Service 2004). Monte Carlo Simulation T h e values for the l ifespan o f ghost f i sh ing traps i n this case study was an average o f thirteen other case studies, some o f w h i c h had asymmetr ica l variances. In the cases where variances were asymmetr ica l , the average o f the variances i n each case study was taken and used towards the f o l l o w i n g formula : S d ( ( X + Y ) / 1 3 ) = (sqr t (Var (X) + V a r ( Y ) ) ) / 1 3 resul t ing i n a mean o f 0.417 + 141%/-100% years (0.417 ± 141%, w i t h the lower l i m i t truncated at 0). The morta l i ty rate is taken as the mean f rom three values, 53.8 crabs- t r a p " ' y e a r ' ± 3 4 . 1 % , as expla ined i n the narrative above. The dis t r ibut ion o f ghost f ishing mor ta l i ty as a percentage o f the reported landed catch is shown i n F igure 3.1.8. Louisiana blue crab fishery 1988 Parameter list Parameter i : 2,800 fishers Parameter i i : 200 traps Parameter 1: 560,000 traps 26 X Parameter 2: 10% - Parameter 3: 56,000 traps X Parameter 4: 100% = Parameter 5: 65,000 traps X Parameter 6: 0.417 + 141%/-100% years Parameter 7: 23,366 traps / Parameter 8: 25.8 crabs- trap"'year' 1 = Parameter 9: 502,835 crabs X Parameter 10: 0.16 k g = Parameter 11: 95,706 kg •H- Parameter 12: 24,366,026 k g = Resul t : 0.36% Narrative A p p r o x i m a t e l y 2,800 l icensed c o m m e r c i a l blue crab (Callinectes sapidus) fishers used an average o f 200 traps each i n L o u i s i a n a i n 1988 (Roberts and T h o m p s o n 1982 ci ted i n G u i l l o r y 1993). G u i l l o r y (1999) estimates an annual trap loss rate o f 10% i n the fishery; 56,000 traps w o u l d be added to the ghost f i sh ing fleet each year. In the same paper, he assumes that a l l these lost traps are ghost f i shing, but gives no estimate for the durat ion o f their functional l ives , so the average (0.417 years) was taken f rom other case studies where this parameter was specif ied. The values f rom those case studies were: 2.2 years i n the 1984 B C Dungeness crab fishery (Breen 1987); 0.418 years i n the 1969-70 A l a s k a red k i n g crab fishery ( K i m k e r 1990); 0.375 years ± 1 1 % i n the 1986 C a l i f o r n i a Dungeness crab fishery (Kennedy 1986); 0.245 +44%/-19% years i n the 1992 C o o k Inlet red k i n g crab fishery ( K i m k e r 1990); and 0.243 ± 3 3 % years. The last value is a composi te i n itself, obtained f rom 0.241 +20.8%/-27.1% years ( K i m k e r 1990) and 0.245 +44%/-19% years ( K i m k e r 1990), and used i n 9 case studies: 1993 B e r i n g Sea crab fishery; 1993 B e r i n g Sea red k i n g crab fishery; 1990 B r i s t o l B a y red k i n g crab fishery; 1991 B r i s t o l B a y red k i n g crab fishery; 1994 B r i s t o l B a y red k i n g crab fishery; 1996 B r i s t o l B a y red k i n g crab fishery; C h i n i a k B a y red k i n g crab fishery i n the 1990s; and 2 1991 eastern B e r i n g Sea snow crab fisheries. A s s u m i n g that the rate o f entry o f lost traps was equal to the rate o f degradation, 23,366 traps are estimated to be ghost f i sh ing i n the area. The study o f ghost f ishing by blue crab traps that corresponds to the f i shing effort used above estimates an average o f 25.8 ind iv idua l s k i l l e d per trap. A p p l i e d to the previous parameter, the annual mortal i ty f rom ghost traps is 602,835 crabs. The average weight o f b lue crabs i n October 1999 to N o v e m b e r 2000 landings was 0.35 lbs, or 0.16 k g ( M u r p h y et al. 2001) . The weight o f 27 ghost f ishing catch is thus 95,706 k g , or 0.36% o f the reported 24,366,026 k g landed i n 1988 (Nat iona l M a r i n e Fisher ies Service 2004). Monte Carlo Simulation The values for the l i fespan o f ghost f ishing traps i n this case study was an average o f thirteen other case studies, some o f w h i c h had asymmetr ica l variances. In the cases where variances were asymmetr ica l , the average o f the variances i n each case study was taken and used towards the f o l l o w i n g formula: S d ( ( X + Y ) / 1 3 ) = ( sq r t (Var (X) + V a r ( Y ) ) ) / 1 3 resul t ing i n a mean o f 0.417 + 141%/-100% years (0.417 ± 1 4 1 % , w i t h the lower l imi t truncated at 0). The dis tr ibut ion o f ghost f i sh ing mor ta l i ty as a percentage o f the reported landed catch is shown i n F igure 3.1.9. Newfoundland snow crab fishery 1974 Parameter list Parameter i : N o t p rov ided Parameter i i : N o t p rov ided Parameter 1: 12,000 traps Parameter 2: N o t p rov ided Parameter 3: 1,000 traps lost per year X Parameter 4: 100% = Parameter 5: 1-000 traps X Parameter 6: 0.417 + 141%/-100% years = Parameter 7: 417 traps X Parameter 8: 10.4 ± 5 0 % crabs- trap"1 year = Parameter 9: 4,339 crabs X Parameter 10: 0.57 k g = Parameter 11: 2,490 kg -f- Parameter 12: 33,000,000 k g = Resul t : 0.08% Narrative A n estimated 12,000 traps were fished i n 1974 i n the N e w f o u n d l a n d snow crab (Chionoecetes opilio) f ishery, 1,000 o f w h i c h were lost and assumed to be capable o f 28 ghost f ishing ( M i l l e r 1977). In the absence o f an estimate for the functional l i fe o f the ghost traps, an average o f ghost f ishing per iod (0.417 years) was taken f rom a l l other case studies that had a value for this parameter was used. The values from those case studies were: 2.2 years i n the 1984 B C Dungeness crab fishery (Breen 1987); 0.418 years i n the 1969-70 A l a s k a red k i n g crab fishery ( K i m k e r 1990); 0.375 years ± 1 1 % i n the 1986 C a l i f o r n i a Dungeness crab f i shery . (Kennedy 1986); 0.245 +44%/-19% years i n the 1992 C o o k Inlet red k i n g crab f ishery ( K i m k e r 1990); and 0.243 ± 3 3 % years. The last va lue is a composi te i n itself, obtained f rom 0.241 +20.8%/-27.1% years ( K i m k e r 1990) and 0.245 +44%/-19% years ( K i m k e r 1990), and used i n 9 case studies: 1993 B e r i n g Sea crab fishery; 1993 B e r i n g Sea red k i n g crab fishery; 1990 B r i s t o l B a y red k i n g crab fishery; 1991 B r i s t o l B a y red k i n g crab fishery; 1994 B r i s t o l B a y red k i n g crab fishery; 1996 B r i s t o l B a y red k i n g crab fishery; C h i n i a k B a y red k i n g crab fishery i n the 1990s; and 2 1991 eastern B e r i n g Sea snow crab fisheries. A s s u m i n g that the rate o f entry o f lost traps was equal to the rate o f degradation, 417 traps are estimated to be ghost f ishing i n the area annual ly . In an experiment to determine whether traps continue to f ish after bait is gone, ( M i l l e r 1977) d iscovered that traps w i t h dead crabs as bait attracted 0.6 crabs per trap, none o f w h i c h escaped. The study took place over 3 weeks , so assuming that the catch rate remained constant throughout the year, the annual ghost f ishing mor ta l i ty f rom a l l traps is 4,339 ind iv idua l s . W i t h the average weight o f snow crabs at 0.57 k g (Industry Canada 2004), the weigh t o f ghost catches f rom traps w i t h dead crabs as bait is 2 ,490 k g , or 0 .75% o f the reported 33,000,000 k g landed i n N e w f o u n d l a n d i n 1987 ( M i l l e r 1977). T h i s is higher than the 0 .30% that M i l l e r (1977) estimated. F o l l o w i n g the same calculat ions, traps us ing squid as bait attract 537 ind iv idua l s per trap per year or 31 ind iv idua l s per trap i n 3 weeks , resul t ing i n a ghost f i sh ing catch o f 3.9% o f the reported landings. Traps w i t h a combina t ion o f squid and dead crab as bait attracted 132 ind iv idua l s per trap per year or 7.6 ind iv idua ls per trap i n 3 weeks ( M i l l e r 1977), resul t ing i n a ghost f ishing catch o f 0 .97% o f the report ing landings, and traps w i t h no bait attract 5.2 ind iv idua ls per trap per year or 0.3 indiv iduals per trap i n 3 weeks ( M i l l e r 1977), resul t ing i n a ghost f i sh ing catch o f 0 .08% o f the reported landings. Monte Carlo Simulation M i l l e r (1977) estimated the mortal i ty rate o f snow crab b y lost traps i n N e w f o u n d l a n d to be 0.6 ± 0 . 3 crabs per trap i n 3 weeks . A s s u m i n g a constant mor ta l i ty rate, this is 10.4 ± 5 0 % crabs- trap^year" 1. The values for the l i fespan o f ghost f i sh ing traps i n this case study was an average o f thirteen other case studies, some o f w h i c h had asymmetr ica l variances. In the cases where variances were asymmetr ica l , the average o f the variances i n each case study was taken and used towards the f o l l o w i n g formula: 29 S d ( ( X + Y ) / 1 3 ) = ( sqr t (Var (X) + V a r ( Y ) ) ) / 1 3 resul t ing i n a mean o f 0.417 + 141%/-100% years (0.417 ± 141%, w i t h the lower l imi t truncated at 0). The average weight o f snow crabs is the mean o f two snow crab case studies i n the G u l f o f St. Lawrence . V i e n n e a u and M o r i y a s u (1994) estimated 0.572 k g and M a l l e t at al. (1998) estimated 0.574 k g . The mean is therefore 0.574 k g ± 0 .35%. The dis t r ibut ion o f ghost f ishing mor ta l i ty as a percentage o f the reported landed catch is shown i n F igu re 3.1.10. Pacific Coast Dungeness crab fishery 1971-1972 Parameter list Parameter i : N o t p rov ided Parameter i i : N o t p rov ided Parameter 1: 138,000 traps X Parameter 2: 10% = Parameter 3: 13,800 traps X Parameter 4: 9 6 . 9 % + 3.1%/- 5 1 . 3 % = Parameter 5: 13,366 traps X Parameter 6: 0.417 + 141%/-100% years = Parameter 7: 5,577 traps X Parameter 8: 22.0 ± 4 2 % crabs- trap"'year - Parameter 9: 122,569 crabs X Parameter 10: 0.68 k g = Parameter 11: 83,347 kg Parameter 12: 3,538,080 k g Resul t : 2.4% Narrative A p p r o x i m a t e l y 138,000 Dungeness crab (Cancer magister) pots were used i n Wash ing ton , Oregon , and Ca l i fo rn i a i n the 1971-1972 season, o f w h i c h 10% was lost (Pacif ic M a r i n e Fisher ies C o m m i s s i o n 1978); therefore, 13,800 traps were lost that year. There was no ind ica t ion o f whether a l l the traps were ghost f ishing. T o achieve a conservative estimate, it was assumed that not a l l the lost traps w o u l d ghost f ish, and a 30 modif ie r o f 9 6 . 9 % was applied. T h i s modifer was the mean calculated from other case studies where this parameter was specif ied: 4 6 . 5 % i n the 1987 B C Dungeness crab fishery (Breen 1987); 100% i n the 1990 C h i n i a k B a y red k i n g crab fishery (assumed f rom R . M e y e r , pers c o m m to S m o l o w i t z 1978a); 100% i n the 1976 At l an t i c offshore lobster fishery (assumed from S m o l o w i t z 1978b); 100% i n the 1969-1989 G u l f o f St. Lawrence snow crab fishery (assumed f rom V i e n n e a u and M o r i y a s u 1994); 1 0 0 % ' i n the 1987 G u l f o f St. L a w r e n c e snow crab fishery ( M a l l e t et al. 1998); 100% i n the 1987 N e w B r u n s w i c k snow crab fishery (Ma l l e t et al. 1998); 100% i n the L o u i s i a n a b lue crab fishery c. 1993 ( G u i l l o r y 1999); 100% i n the 1988 L o u i s i a n a b lue crab fishery ( G u i l l o r y 1999); 100% i n the 1993 L o u i s i a n a b lue crab fishery ( G u i l l o r y 1999); and 100% i n the 1974 N e w f o u n d l a n d snow crab fishery ( M i l l e r 1977). The result w a s l 3 , 3 6 6 ghost f ish traps. There was no indica t ion o f h o w long the lost traps w o u l d remain functional , so an average length o f t ime, 0.417 years, was taken from a l l case studies i n w h i c h ghost f i sh ing per iod was specified. The values f rom those case studies were: 2.2 years i n the 1984 B C Dungeness crab fishery (Breen 1987); 0.418 years i n the 1969-70 A l a s k a red k i n g crab fishery ( K i m k e r 1990); 0.375 years ± 1 1 % i n the 1986 C a l i f o r n i a Dungeness crab fishery (Kennedy 1986); 0.245 +44%/-19% years i n the 1992 C o o k Inlet red k i n g crab fishery ( K i m k e r 1990); and 0.243 ± 3 3 % years. The last value is a composi te i n itself, obtained f rom 0.241 +20.8%/-27.1% years ( K i m k e r 1990) and 0.245 +44%/-19% years ( K i m k e r 1990), and used i n 9 case studies: 1993 B e r i n g Sea crab fishery; 1993 B e r i n g Sea red k i n g crab fishery; 1990 B r i s t o l B a y red k i n g crab fishery; 1991 B r i s t o l B a y red k i n g crab fishery; 1994 B r i s t o l B a y red k i n g crab fishery; 1996 B r i s t o l B a y red k i n g crab fishery; C h i n i a k B a y red k i n g crab fishery i n the 1990s; and 2 1991 eastern B e r i n g Sea snow crab fisheries. A s s u m i n g that the rate o f entry o f lost traps was equal to the rate o f degradation, and that the numbers remained constant since then, 5,577 traps are estimated to be ghost f i sh ing inshore i n the area annual ly . D e m o r y (1971 ci ted i n Pac i f i c M a r i n e Fisher ies C o m m i s s i o n 1978) retrieved 117 crab traps i n Oregon, where each trap he ld an average o f 31.1 crabs. Dah l s t rom (1975 c i ted i n Pac i f i c M a r i n e Fisheries C o m m i s s i o n 1978) retr ieved 90 crabs i n 7 crab traps i n Ca l i f o rn i a ; the average catch rate is 12.8 crabs per trap. The length o f t ime i n w h i c h these traps had been ghost f i sh ing is u n k n o w n ; for the purposes o f this case study, it is assumed that they had been f ishing for one year and that the number o f caught crabs was i n equ i l i b r ium - that is, that is the number o f crabs taken f rom the avai lable popula t ion at any g iven t ime. The average o f these catch rates (22.0 individuals- trap"'year" 1), appl ied to a l l lost traps, results i n 122,569 ind iv idua l s k i l l e d by ghost f i sh ing traps per year. W i t h the mean weight per crab at 0.68 k g (Breen 1987), the weight o f ghost f i sh ing catch is 83,347 k g , 2 .4% o f the reported 3,538,080 k g i n landings i n 1974 (Pac i f ic M a r i n e Fisheries C o m m i s s i o n 1978). 31 Monte Carlo Simulation The percentage o f traps lost to ghost f i sh ing is the grand mean calculated from case studies where the parameters were g iven , w h i c h results i n 9 6 . 9 % + 3 .1%/- 51 .3% o f lost traps (96 .9% ± 51 .3%, w i t h the upper l imi t truncated at 100%). The values for the l i fespan o f ghost f i sh ing traps i n this case study was an average o f thirteen other case studies, some o f w h i c h had asymmetr ica l variances. In the cases where variances were asymmetr ica l , the average o f the variances i n each case study was taken and used towards the f o l l o w i n g formula : S d ( ( X + Y ) / 1 3 ) = ( sqr t (Var (X) + V a r ( Y ) ) ) / 1 3 resul t ing i n a mean o f 0.417 + 141%/-100% years (0.417 ± 141%, w i t h the lower l imi t truncated at 0). M o r t a l i t y rate was averaged f rom other Pac i f i c Coast mor ta l i ty rates. D e m o r y (1971 ci ted i n Pac i f i c M a r i n e Fisher ies C o m m i s s i o n 1978) estimated an average o f 31.1 crabs k i l l e d per trap and D a h l s t r o m (1975 ci ted i n Pac i f i c M a r i n e Fisher ies C o m m i s s i o n 1978), estimated an average o f 12.8 crabs k i l l e d per trap, so the average morta l i ty is 22.0 ± 4 2 % . The dis tr ibut ion o f ghost f i sh ing mor ta l i ty as a percentage o f the reported landed catch is shown i n F igure 3.1.11. Washington State Dungeness crab fishery 1972 Parameter list Parameter i : 206 - not used i n calculat ions Parameter i i : N o t p rov ided Parameter 1: 54,518,traps X Parameter 2: 15.7% — Parameter 3: 8,559 traps X Parameter 4: 100% = Parameter 5: 8,559 traps X Parameter 6: 0 . 417+ 141%/-100% years = Parameter 7: 3,571 traps X Parameter 8: 114 crabs- trap _ 1year"' 32 X Parameter 9: Parameter 10 Parameter 11: Parameter 12 Resul t : 407,355 crabs 0.68 k g 277,001 kg 5,616,566 4.9% Narrative In 1972, 149 boats f ished 40,518 traps for Dungeness crabs (Cancer magister) i n Wash ing ton State (Tegelberg 1974). A n addi t ional 57 out o f state boats fished an addi t ional 14,000 traps, b r ing ing the total number o f traps used i n the area to 54,518, o f w h i c h 15.7%, or 8,559 traps, were lost (Tegelberg 1974). A n addi t ional 1.5% o f the traps were discarded as unusable at the end o f the season; since the estimate o f trap loss is separate from those that were unusable, it is assumed that a l l the lost traps were capable o f ghost f ishing. N o estimate for the funct ional l ifespan o f the traps was avai lable , so an average length o f t ime, 0.417 years, was taken f rom a l l case studies i n w h i c h ghost f ishing per iod was specified. The values f rom those case studies were: 2.2 years i n the 1984 B C Dungeness crab fishery (Breen 1987); 0.418 years i n the 1969-70 A l a s k a red k i n g crab fishery ( K i m k e r 1990); 0.375 years ± 1 1 % i n the 1986 C a l i f o r n i a Dungeness crab fishery ( K e n n e d y 1986); 0.245 +44%/-19% years i n the 1992 C o o k Inlet red k i n g crab fishery ( K i m k e r 1990); and 0.243 ± 3 3 % years. The last va lue is a composi te i n itself, obtained f rom 0.241 +20.8%/-27.1% years ( K i m k e r 1990) and 0.245 +44%/-19% years ( K i m k e r 1990), and used i n 9 case studies: 1993 B e r i n g Sea crab fishery; 1993 B e r i n g Sea red k i n g crab fishery; 1990 B r i s t o l B a y red k i n g crab fishery; 1991 B r i s t o l B a y red k i n g crab fishery; 1994 B r i s t o l B a y red k i n g crab fishery; 1996 B r i s t o l B a y red k i n g crab fishery; C h i n i a k B a y red k i n g crab f ishery i n the 1990s; and 2 1991 eastern B e r i n g Sea snow crab fisheries. A s s u m i n g that the rate o f entry o f lost traps was equal to the rate o f degradation and that the values from the previous parameters have remained constant since, 3,571 traps are estimated to be ghost f ishing inshore i n the area annually. In a 1982 study i n the C o l u m b i a R i v e r estuary, M u i r et al. (1984) used four standard commerc ia l crab traps to examine the escape o f Dungeness crabs i n s imulated lost traps. O v e r a 28-day per iod , 35 crabs died i n the four traps (an average o f 8.75 crabs per trap). A s s u m i n g the same morta l i ty rate throughout the year, the year ly morta l i ty can be extrapolated to 114 ind iv idua l s per trap. The total annual mor ta l i ty from a l l lost traps is thus 407,355. T o express this mor ta l i ty as a percentage o f the reported landings, the average weight for Dungeness crab (0.68 k g ; B reen 1987) was appl ied. The weight o f Dungeness crab morta l i ty f rom ghost traps was therefore estimated to be 277,001 k g , or 33 4 .9% o f the reported 5,616,566 k g landed i n Wash ing ton State i n 1972 (Nat iona l M a r i n e Fisheries Service 2004) . Monte Carlo Simulation The values for the l ifespan o f ghost f i sh ing traps i n this case study was an average o f thirteen other case studies, some o f w h i c h had asymmetr ica l variances. In the cases where variances were asymmetr ica l , the average o f the variances i n each case study was taken and used towards the f o l l o w i n g formula: S d ( ( X + Y ) / 1 3 ) = ( sqr t (Var (X) + V a r ( Y ) ) ) / 1 3 resul t ing i n a mean o f 0.417 + 141%/-100% years (0.417 ± 141%, w i t h the l ower l imi t truncated at 0). The dis t r ibut ion o f ghost f ishing mor ta l i ty as a percentage o f the reported landed catch is shown i n F igure 3.1.12. Washington State Dungeness crab fishery 1973 Parameter list Parameter i : 147 vessels not used i n calculat ions Parameter i i : N o t p rov ided Parameter 1: 37,637 traps X Parameter 2: 15.7% = Parameter 3: 5,909 traps X Parameter 4: 100% = Parameter 5: 5,909 traps X Parameter 6: 0 . 4 1 7 + 141 % / - 1 0 0 % years = Parameter 7: 2,466 traps X Parameter 8: 114 crabs- trap"'year"' — Parameter 9: 281,221 crabs X Parameter 10: 0.68 k g - Parameter 11: 191,230 kg -r- Parameter 12: 2,074,041 k g = Result : 9.2% Narrative In 1973, 147 boats fished 37,637 Dungeness crab (Cancer magister) traps i n Wash ing ton State, o f w h i c h 15.7%, or 5,909 traps, were lost (Tegelberg 1974). A n 34 addi t ional 1.5% o f the traps were discarded as unusable at the end o f the season; since the estimate o f trap loss is separate from those that were unusable, it is assumed that a l l the lost traps were capable o f ghost f ishing. N o estimate for the functional l ifespan o f the traps was avai lable , so an average length o f t ime, 0.417 years, was taken f rom a l l case studies i n w h i c h ghost f i sh ing per iod was specified. The values from those case studies were: 2.2 years i n the 1984 B C Dungeness crab fishery (Breen 1987); 0.418 years i n the 1969-70 A l a s k a red k i n g crab fishery ( K i m k e r 1990); 0.375 years ± 1 1 % i n the 1986 C a l i f o r n i a Dungeness crab fishery (Kennedy 1986); 0.245 +44%/-19% years i n the 1992 C o o k Inlet red k i n g crab fishery ( K i m k e r 1990); and 0.243 ± 3 3 % years. The last va lue is a composi te i n itself, obtained from 0.241 +20.8%/-27.1% years ( K i m k e r 1990) and 0.245 +44%/-19% years ( K i m k e r 1990), and used i n 9 case studies: 1993 B e r i n g Sea crab fishery; 1993 B e r i n g Sea red k i n g crab fishery; 1990 B r i s t o l B a y red k i n g crab fishery; 1991 B r i s t o l B a y red k i n g crab fishery; 1994 B r i s t o l B a y red k i n g crab fishery; 1996 B r i s t o l B a y red k i n g crab fishery; C h i n i a k B a y red k i n g crab fishery i n the 1990s; and 2 1991 eastern B e r i n g Sea snow crab fisheries. A s s u m i n g that the rate o f entry o f lost traps was equal to the rate o f degradation and that the values f rom the previous parameters have remained constant since, 2,466 traps are estimated to be ghost f i sh ing inshore i n the area annual ly . In a 1982 study i n the C o l u m b i a R i v e r estuary, M u i r et al. (1984) used four standard c o m m e r c i a l crab traps to examine the escape o f Dungeness crabs i n s imulated lost traps. O v e r a 28-day per iod, 35 crabs d ied i n the four traps (an average o f 8.75 crabs per trap). A s s u m i n g the same morta l i ty rate throughout the year, the year ly mor ta l i ty can be extrapolated to 114 ind iv idua ls per trap. The total annual morta l i ty f rom a l l lost traps is thus 281 ,221. T o express this morta l i ty as a percentage o f the reported landings, the average weight for Dungeness crab (0.68 k g ; B r e e n 1987) was appl ied. The weight o f Dungeness crab morta l i ty from ghost traps was therefore estimated to be 191,230 k g , or 9 .2% o f the reported 2,074,041 k g landed i n W a s h i n g t o n State i n 1973 (Nat iona l M a r i n e Fisheries Service 2004). Monte Carlo Simulation The values for the lifespan o f ghost f i sh ing traps i n this case study was an average o f thirteen other case studies, some o f w h i c h had asymmetr ica l variances. In the cases where variances were asymmetr ica l , the average o f the variances i n each case study was taken and used towards the f o l l o w i n g formula : S d ( ( X + Y ) / 1 3 ) = ( sqr t (Var (X) + V a r ( Y ) ) ) / 1 3 35 resul t ing i n a mean o f 0.417 + 141%/-100% years (0.417 ± 141%, w i t h the lower l imi t truncated at 0). The dis tr ibut ion o f ghost f i sh ing mor ta l i ty as a percentage o f the reported landed catch is shown i n F igure 3.1.13. C a s e s tudies w i t h 3 m i s s i n g p a r a m e t e r s : Alabama blue crab fishery 1999-2000 Parameter list Parameter i : 174 fishers X Parameter i i : 150 traps per fisher = Parameter 1: 26,100 traps X Parameter 2: 3 5 % ± 4 3 % = Parameter 3: 9,135 traps X Parameter 4: 9 6 . 9 % + 3 .1%/- 51 .3% = Parameter 5: 8,848 traps X Parameter 6: 0.417 + 141%/-100% years = Parameter 7: 3,692 traps X Parameter 8: 53.8 ± 3 4 . 1 % crabs- trap"'year = Parameter 9: 198,607 crabs X Parameter 10: 0.16 k g = Parameter 11: 31,531 kg -f- Parameter 12: 1,100,558 k g = Resul t : 2.9% Narrative In the 1999-2000 season, there were 174 l icensed fishers ( G u i l l o r y et al. 2001) w i t h 150 traps each (Heath 1998 ci ted i n G u i l l o r y et al. 2001) i n the A l a b a m a blue crab (Callinectes sapidus) fishery, so there were 26,100 traps i n use that season i n total. L . Ha r tman (pers. c o m m . to G u i l l o r y et al. 2001) estimates the annual rate o f trap loss, i nc lud ing theft, i n the fishery to be between 20 and 50%. I f the midrange is used as the annual rate o f trap loss, the calculated number o f traps lost that year is 9,135. Because there is no estimation o f what proport ion o f these traps was stolen and therefore not ghost f ishing, an average (96.9%) is taken o f the values f rom other case studies where such a number is g iven : 4 6 . 5 % i n the 1987 B C Dungeness crab fishery (Breen 1987); 100% i n the 1990 C h i n i a k B a y red k i n g crab fishery (assumed f rom R . 36 M e y e r , pers c o m m to S m o l o w i t z 1978a); 100% i n the 1976 At l an t i c offshore lobster f ishery (assumed from S m o l o w i t z 1978b); 100% i n the 1969-1989 G u l f o f St. Lawrence snow crab fishery (assumed f rom V i e n n e a u and M o r i y a s u 1994); 100% i n the 1987 G u l f o f St. Lawrence snow crab fishery ( M a l l e t et al. 1998); 100% i n the 1987 N e w B r u n s w i c k snow crab fishery ( M a l l e t et al. 1998); 100% i n the L o u i s i a n a blue crab f ishery c. 1993 ( G u i l l o r y 1999); 100% i n the 1988 L o u i s i a n a b lue crab fishery ( G u i l l o r y 1999); 100% i n the 1993 L o u i s i a n a b lue crab fishery ( G u i l l o r y 1999); and 100% i n the 1974 N e w f o u n d l a n d snow crab fishery ( M i l l e r 1977). The result was96,853 ghost f ish traps. W h e n that is appl ied, the number o f traps lost to ghost f i sh ing is calculated to be 8,848. The same method is used for length o f t ime a lost trap w o u l d be ac t ive ly ghost f i sh ing (0.417 years). The values f rom those case studies were: 2.2 years i n the 1984 B C Dungeness crab fishery (Breen 1987); 0.418 years i n the 1969-70 A l a s k a red k i n g crab f ishery ( K i m k e r 1990); 0.375 years ± 1 1 % i n the 1986 C a l i f o r n i a Dungeness crab fishery ( K e n n e d y 1986); 0.245 +44%/-19% years i n the 1992 C o o k Inlet red k i n g crab fishery ( K i m k e r 1990); and 0.243 ± 3 3 % years. The last value is a composi te i n itself, obtained f rom 0.241 +20.8%/-27.1% years ( K i m k e r 1990) and 0.245 +44%/-19% years ( K i m k e r 1990), and used i n 9 case studies: 1993 B e r i n g Sea crab fishery; 1993 B e r i n g Sea red k i n g crab fishery; 1990 B r i s t o l B a y red k i n g crab fishery; 1991 B r i s t o l B a y red k i n g crab fishery; 1994 B r i s t o l B a y red k i n g crab fishery; 1996 B r i s t o l B a y red k i n g crab fishery; C h i n i a k B a y red k i n g crab fishery i n the 1990s; and 2 1991 eastern B e r i n g Sea snow crab fisheries. The number o f traps ghost f i sh ing is calculated to be 3,692, assuming the system is at equ i l ib r ium. There is no estimate o f morta l i ty per trap i n that fishery, so the average o f mortal i t ies (53.8 crabs per pot per year) i n other b lue crab fisheries is applied. The number o f crabs k i l l e d b y ghost traps is therefore 198,607. A generic estimate o f the average weight o f blue crabs (0.16 kg) was obtained ( M u r p h y et al. 2001) ; the weight o f blue crab morta l i ty from ghost traps is therefore estimated to be 31,531 k g , w h i c h is 2 .9% o f the 1,100,558 k g landed i n A l a b a m a i n 2001 (Nat iona l M a r i n e Fisher ies Service 2004). Monte Carlo simulation Gear loss i n this fishery was estimated to be between 2 0 % and 5 0 % ( L . Har tman pers. c o m m . to G u i l l o r y etal. 2001) , w h i c h equates to a midpo in t o f 3 5 % ± 4 3 % . T h e percentage o f traps lost to ghost f ishing is the grand mean calculated from case studies where the parameters were g iven , w h i c h results i n 9 6 . 9 % + 3 .1%/- 51 .3% o f lost traps (96 .9% ± 51 .3%, w i t h the upper l imi t truncated at 100%). The values for the l i fespan o f ghost f ishing traps i n this case study was an average o f thirteen other case studies, some o f w h i c h had asymmetr ica l variances. In the cases 37 where variances were asymmetr ica l , the average o f the variances i n each case study was taken and used towards the f o l l o w i n g formula: S d ( ( X + Y ) / 1 3 ) = ( sq r t (Var (X) + V a r ( Y ) ) ) / 1 3 resul t ing i n a mean o f 0.417 + 141%/-100% years (0.417 ± 141%, w i t h the lower l imi t truncated at 0). The mortal i ty rate is taken as the mean from three values, 53.8 crabs-trap^year" 1 ± 3 4 . 1 % , as expla ined i n the narrative above. The dis t r ibut ion o f ghost f i sh ing mor ta l i ty as a percentage o f the reported landed catch is shown i n F igure 3.1.14. Eastern Bering Sea crab fishery 1993 Parameter list Parameter i : N o t p rov ided Parameter i i : N o t p rov ided Parameter 1: 71,000 traps X Parameter 2: 2 0 % = Parameter 3: 14,200 traps X Parameter 4: 9 6 . 9 % + 3 . 1 % / - 5 1 . 3 % = Parameter 5: 13,753 traps X Parameter 6: 0.243 ± 3 3 % years = Parameter 7: 3,340 traps X Parameter 8: 9.03 ± 47 .4% crabs- trap"'year = Parameter 9: 30,176 crabs X Parameter 10: 1.16 ± 2 1 . 8 % k g = Parameter 11: 42,269 kg Parameter 12: 63,282,000 k g = Resul t : 0.07% Narrative In the eastern B e r i n g Sea crab fishery, 71,000 traps were registered for use i n 1993 ( D . Tracy , pers. c o m m . to Stevens 1996), though the target species was not identif ied. T o provide a conservat ive estimate, it w i l l be assumed that the registered traps were for a l l crab species. A n annual trap loss rate o f 2 0 % ( W . Nippes , pers. c o m m . to 38 K r u s e and K i m k e r 1993) for the B e r i n g S t ra i t /Aleu t ian Islands region red k i n g crab (Paralithocles camtschatica) is applied i n the absence o f a loss rate for the 71,000 traps, result ing in 14,200 traps lost. There was no indica t ion o f whether a l l the traps were ghost f ishing. T o achieve a conservative estimate, it was assumed that not a l l the lost traps w o u l d ghost f ish, and a modi f ie r o f 96 .9% was appl ied. Th i s modifer was the mean calculated from other case studies where this parameter was specif ied: 4 6 . 5 % i n the 1987 B C Dungeness crab fishery (Breen 1987); 100% i n the 1990 C h i n i a k B a y red k i n g crab fishery (assumed f rom R . M e y e r , pers c o m m to S m o l o w i t z 1978a); 100% i n the 1976 At l an t i c offshore lobster fishery (assumed f rom S m o l o w i t z 1978b); 100% i n the 1969-1989 G u l f o f St. L a w r e n c e snow crab fishery (assumed f rom V i e n n e a u and M o r i y a s u 1994); 100% i n the 1987 G u l f o f St. Lawrence snow crab fishery (Ma l l e t et al. 1998); 100% i n the 1987 N e w B r u n s w i c k snow crab fishery (Ma l l e t et al. 1998); 100% i n the L o u i s i a n a blue crab fishery c. 1993 ( G u i l l o r y 1999); 100% i n the 1988 L o u i s i a n a b lue crab fishery ( G u i l l o r y 1999); 100% i n the 1993 L o u i s i a n a blue crab fishery ( G u i l l o r y 1999); and 100% i n the 1974 N e w f o u n d l a n d snow crab fishery ( M i l l e r 1977). T h e result w a s l 3 , 7 5 3 ghost f i sh ing traps. S ince A l a s k a n legis la t ion requires that a l l crab fisheries (aside f rom Dungeness crab fisheries) use 30 thread cotton twine (Kruse and K i m k e r 1993), the results for 30 thread twine f rom K i m k e r ' s (1990) study o f var ious sizes biodegradable twine was consulted to estimate h o w long these traps w o u l d r emain capable o f ghost f ishing. E i g h t repetitions i n C o o k Inlet y ie lded an average o f 89.4 days (wi th a range o f 50-106 days), or 0.245 years, to total twine degradation, and f ive repetitions i n Pr ince W i l l i a m S o u n d y ie lded an average o f 87.9 days (wi th a range o f 76-99 days), or 0.241 years, to total twine degradation. N o experiment was done i n the B e r i n g Sea, so the average (0.243 years) o f the values f rom the two locations was used instead. A s s u m i n g that the system was at equ i l ib r ium, whe re in ghost traps were added at the same rate as they were be ing rendered inoperable, 3,340 traps are calculated to be ghost f ishing annually. Because the target crab species was not ident i f ied for this case study, an average o f crab mortal i t ies (9.03 crabs- trap^'year" 1) i n other case studies situated i n A l a s k a was appl ied to calculate that 30,176 "crabs" are k i l l e d per year by ghost traps. The mortal i t ies from the case studies are: 1.54 tanner crabs per trap i n K o d i a k , A l a s k a i n the 1990s (Stevens et al. 2000); 0.03 red k i n g crabs per trap i n K o d i a k , A l a s k a i n the 1990s (Stevens et al. 2000); 4 "crabs" per trap i n C h i n i a k , A l a s k a i n the 1990s (Stevens 1995); 3.9 k i n g crab per trap i n C h i n i a k , A l a s k a i n the 1990s (Stevens 1996); and 35.7 tanner crabs per trap i n C h i n i a k , A l a s k a i n the 1990s (Stevens 1996). 39 A n average was also taken from weight estimates o f c o m m e r c i a l l y targeted crabs i n A l a s k a : snow crab Chionoecetes opilio (0.16 k g ; M a l l e t et al. 1998); red k i n g crab (6.5 lbs, or 2.95 k g ; K r u s e and K i m k e r 1993); and Dungeness crab (0.68 k g ; B r e e n 1987). The weight o f crab morta l i ty was therefore estimated to be 42,269 k g , w h i c h is 0 .07% o f the 63,282,000 k g o f crab landed i n A l a s k a i n 1997 (Reeves and Turnock 1999). 1997 landing informat ion was used because 1993 landings for the B e r i n g Strait were not available. Monte Carlo simulation The percentage o f traps lost to ghost f i sh ing is the grand mean calculated from, case studies where the parameters were g iven , w h i c h results i n 96 .9% + 3 .1%/ - 5 1 . 3 % o f lost traps (96 .9% ± 51 .3%, w i t h the upper l imi t truncated at 100%). The value for the l i fespan o f ghost f i sh ing traps i n this case study was an average o f two other values that had asymmetr ica l variances (0.241 years w i t h a range o f 0.208 to 0.271 years and 0.245 years w i t h a range o f 0.189 to 0.290 years). A n average was thus taken o f the variances for each case study, and the variance o f the mean was obtained f rom the f o l l o w i n g formula : S d ( ( X + Y ) / 2 ) = l / 2 s q r t ( V a r ( X ) + V a r ( Y ) ) result ing i n a mean o f 0.243 years ± 3 3 % . The morta l i ty rate for this case study is taken from a l l other A l a s k a n case studies. K i m k e r (1990) estimated a mortal i ty o f 3.9 red k i n g crabs- trap"'year" 1 and 35.7 tanner crabs i n traps w i t h 30 twine thread, Stevens et al. (2000) estimated 0.03 red k i n g crabs-trap"'year"' and 1.54 tanner crabs- trap"'year"', and Stevens (1996) found 4 unspeci f ied species o f crab per trap per year. The mean mor ta l i ty is therefore 9.03 ± 47.4 crabs- trap* 'year - 1 . The species weight was taken as the mean f rom other A l a s k a n case studies, where snow crab is 0.16 k g ( M a l l e t et al. 1998), red k i n g crab is 2.95 k g (Kruse and K i m k e r 1993) and Dungeness crab is 0.68 k g (Breen 1987). The mean weight is therefore 1.16 ± 21 .8% k g . The dis t r ibut ion o f ghost f ishing mor ta l i ty as a percentage o f the reported landed catch is shown i n F igu re 3.1.15. 40 Eastern Bering Sea snow crab fishery 1999 Parameter list Parameter i : N o t p rov ided Parameter i i : N o t p rov ided Parameter 1: 50,720 traps X Parameter 2: 1% = Parameter 3: 507 traps X Parameter 4: 9 6 . 9 % + 3 .1%/- 5 1 . 3 % Parameter 5: 491 traps Parameter 6: 0.243 ± 3 3 % years = Parameter 7: 119 traps X Parameter 8: 42.7 ± 3 .9% crabs- trap"'year = Parameter 9: 5,089 crabs X Parameter 10: 0.57 ± 0 .35% k g — Parameter 11: 2,490 kg -7" Parameter 12: 8,808,912 k g = Resul t : 0.03% Narrative In 1999, 50,720 traps were registered for the B e r i n g Sea snow crab (Chionoecetes opilio) f ishery (Stevens et al. 2000). The rate o f trap loss is estimated at 1% o f traps used (R. M o r r i s o n pers. c o m m . to Stevens et al. 2000), result ing i n 507 traps lost annual ly . Stevens et al. (2000) indicate that the estimate o f trap loss includes traps that require replac ing , so not a l l lost traps end up ghost f ishing. There was no ind ica t ion o f whether a l l the traps were ghost f ishing. T o achieve a conservative estimate, it was assumed that not a l l the lost traps w o u l d ghost f ish, and a modi f ie r o f 9 6 . 9 % was appl ied. T h i s modifer was the mean calculated f rom other case studies where this parameter was specif ied: 4 6 . 5 % i n the 1987 B C Dungeness crab fishery (Breen 1987); 100% i n the 1990 C h i n i a k B a y red k i n g crab fishery (assumed f rom R . M e y e r , pers c o m m to S m o l o w i t z 1978a); 100% i n the 1976 At l an t i c offshore lobster fishery (assumed f rom S m o l o w i t z 1978b); 100% i n the 1969-1989 G u l f o f St. Lawrence snow crab f ishery (assumed from V i e n n e a u and M o r i y a s u 1994); 100% i n the 1987 G u l f o f St. L a w r e n c e snow crab fishery ( M a l l e t et al. 1998); 100% i n the 1987 N e w B r u n s w i c k snow crab fishery ( M a l l e t et al. 1998); 100% i n the L o u i s i a n a blue crab fishery c. 1993 ( G u i l l o r y 1999); 100% i n the 1988 L o u i s i a n a blue crab fishery ( G u i l l o r y 1999); 100% i n the 1993 L o u i s i a n a blue crab 41 fishery ( G u i l l o r y 1999); and 1 0 0 % i n the 1974 N e w f o u n d l a n d snow crab fishery ( M i l l e r 1977). The result was491 traps lost to ghost f ishing. S ince A l a s k a n legis la t ion requires that a l l crab fisheries (aside from Dungeness crab fisheries) use 30 thread cotton twine (Kruse and K i m k e r 1993), the results for 30 thread twine from K i m k e r ' s (1990) study o f various sizes biodegradable twine was consul ted to estimate h o w long these traps w o u l d remain capable o f ghost f ishing. E i g h t repetitions i n C o o k Inlet y i e lded an average o f 89.4 days (wi th a range o f 50-106 days), or 0.245 years, to total twine degradation, and f i v e repetitions i n Pr ince W i l l i a m Sound y i e lded an average o f 87.9 days (wi th a range o f 76-99 days), or 0.241 years, for total twine degradation. N o experiment was done i n K o d i a k , so the average o f the values (0.243 years) f rom the two locat ions was used instead. A s s u m i n g that the rate o f entry o f lost traps was equal to the rate o f degradation, 119 traps are estimated to be ghost f i sh ing i n K o d i a k annually. S ince there was no estimate o f h o w many snow crabs each trap catches annual ly , the values o f this parameter f rom the St Lawrence snow crab fishery was appl ied instead. V i e n n e a u and M o r i y a s u (1994) estimate the total mor ta l i ty rate o f crab caught b y a lost trap to be 5 0 % o f a funct ioning trap and the average C P U E observed dur ing the spr ing 1993 fishery was 82 ind iv idua l s per trap, so the mor ta l i ty from ghost traps is 41 ind iv idua l s per trap per year. H o w e v e r , i n a t rawl survey o f the G u l f o f St. Lawrence , M a l l e t et al. (1998) found the mean number o f crabs found i n lost traps at 44.3 crabs per trap. The average o f these values (42.7 crabs- trap"'year"') was used i n this case study, resul t ing i n 5,089 crabs lost to ghost traps per year. T o express the number o f crabs caught i n the lost traps as a percentage o f the reported landings (8,808,912 k g ; ( A l a s k a Department o f F i s h and G a m e 2004a) the average weight for red k i n g crab (0.57 kg) was obtained f rom M a l l e t et al. (1998); the weight o f k i n g crab morta l i ty f rom ghost traps was therefore estimated to be 2,920 k g , o r ' 0 .03% o f the reported landings. Monte Carlo simulation The percentage o f traps lost to ghost f ishing is the grand mean calculated f rom case studies where the parameters were g iven , w h i c h results i n 9 6 . 9 % + 3 .1%/- 5 1 . 3 % o f lost traps (96 .9% ± 51 .3%, w i t h the upper l imi t truncated at 100%). The value for the l i fespan o f ghost f i sh ing traps i n this case study was an average o f two other case studies that had asymmetr ica l variances (0.241 years w i t h a range o f 0.208 to 0.271 years and 0.245 years w i t h a range o f 0.189 to 0.290 years). A n average was thus taken o f the variances for each case study, and the variance o f the mean was obtained from the f o l l o w i n g formula : S d ( ( X + Y ) / 2 ) = l / 2 s q r t ( V a r ( X ) + V a r ( Y ) ) 42 resulting in a mean of 0.243 years ± 33%. The mortality rate is the mean of two other case studies involving snow crabs: Stevens et al. (1993) estimated a mortality of 44.3 crabs- trap"1 year"1 and Pecci et al. (1978) estimated a mortality of 41 crabs- trap"'year"1. The mean is thus 42.7 ± 3.9% crabs- trap"'year_1. The average weight of snow crabs is the mean of two snow crab case studies in the Gulf of St. Lawrence. Vienneau and Moriyasu (1994) estimated 0.572 kg and Mallet at al. (1998) estimated 0.576 kg. The mean is therefore 0.574 kg ± 0.35%. The distribution of ghost fishing mortality as a percentage of the reported landed catch is shown in Figure 3.1.16. Florida blue crab fishery 2001 Parameter list Parameter i : 855 fishers Parameter i i : 346 traps per fisher Parameter 1: 295,830 traps Parameter 2: 40% ± 25% Parameter 3: 118,332 traps . Parameter 4: 96.9% + 3.1%/- 51.3% P arameter 5: 114,608 traps ghost fish ing Parameter 6: 0.417 + 141%/-100% years Parameter 7: 47,820 traps Parameter 8: 53.8 ± 34.1% crabs- trap"'year" P arameter 9 : 2,5 72,693 era bs Parameter 10: 0.16 kg Parameter 11: 408,441 kg Parameter 12: 3,379,000 kg Result: 12.1% Narrative In the Florida blue crab (Callinectes sapidus) fishery in 2001, 855 fishers use an average of 346 traps each annually (A. Jackson, pers. comm. to A . Poon, 2004), resulting in a total of 295,830 traps. A . McMillan-Jackson (pers. comm. to Guillory et al. 2001) 43 estimates the trap loss rate i n F l o r i d a to be 30-50%; for the purposes o f this case study, the mean was used, resul t ing i n an estimate o f 118,332 traps lost annually. The estimate o f trap loss inc luded stolen traps, but there was no ind ica t ion o f the proport ion o f traps lost to ghost f ishing. H e n c e , a modi f ie r o f 96 .9% was appl ied. T h i s modifer was the mean calculated f rom other case studies where this parameter was specif ied: 4 6 . 5 % i n the 1987 B C Dungeness crab fishery (Breen 1987); 100% i n the 1990 C h i n i a k B a y red k i n g crab fishery (assumed f rom R . M e y e r , pers c o m m to S m o l o w i t z 1978a); 100% i n the 1976 At lan t i c offshore lobster fishery (assumed f rom S m o l o w i t z 1978b); 100% i n the 1969-1989 G u l f o f St. L a w r e n c e snow crab fishery (assumed f rom V i e n n e a u and M o r i y a s u 1994); 100% i n the 1987 G u l f o f St. L a w r e n c e snow crab fishery ( M a l l e t etal. 1998) ; 100% i n the 1987 N e w B r u n s w i c k snow crab f ishery (Ma l l e t et al. 1998); 100% i n the L o u i s i a n a b lue crab fishery c. 1993 ( G u i l l o r y 1999); 100% i n the 1988 L o u i s i a n a b lue crab fishery ( G u i l l o r y 1999); 100% i n the 1993 L o u i s i a n a b lue crab fishery ( G u i l l o r y 1999) ; and 100% i n the 1974 N e w f o u n d l a n d snow crab fishery ( M i l l e r 1977). The result w a s l 14,608 ghost f ish traps. N o estimate for the functional l ifespan o f the traps was avai lable , so an average length o f t ime, 0.417 years, was taken from a l l case studies i n w h i c h ghost f i sh ing pe r iod was specified. The values from those case studies were: 2.2 years i n the 1984 B C Dungeness crab fishery (Breen 1987); 0.418 years i n the 1969-70 A l a s k a red k i n g crab fishery ( K i m k e r 1990); 0.375 years ± 1 1 % i n the 1986 C a l i f o r n i a Dungeness crab f ishery (Kennedy 1986); 0.245 +44%/-19% years i n the 1992 C o o k Inlet red k i n g crab f ishery ( K i m k e r 1990); and 0.243 ± 3 3 % years. The last value is a composi te i n itself, obtained f rom 0.241 +20.8%/-27.1% years ( K i m k e r 1990) and 0.245 +44%/-19% years ( K i m k e r 1990), and used i n 9 case studies: 1993 B e r i n g Sea crab fishery; 1993 B e r i n g Sea red k i n g crab fishery; 1990 B r i s t o l B a y red k i n g crab fishery; 1991 B r i s t o l B a y red k i n g crab fishery; 1994 B r i s t o l B a y red k i n g crab fishery; 1996 B r i s t o l B a y red k i n g crab fishery; C h i n i a k B a y red k i n g crab fishery i n the 1990s; and 2 1991 eastern B e r i n g Sea snow crab fisheries. A s s u m i n g that the rate o f entry o f lost traps was equal to the rate o f degradation, 47,820 traps are estimated to be ghost f i sh ing i n the area annual ly . In the absence o f morta l i ty rate estimates for the F l o r i d a blue crab fishery, the average o f mor ta l i ty rates f rom L o u i s i a n a blue crab fisheries were appl ied instead. In G u i l l o r y ' s (1993) study o f ghost f ishing by blue crab traps i n Lou i s i ana , an average o f 25.8 indiv iduals d ied per trap. Arcement and G u i l l o r y (1993) performed a compar i son o f morta l i ty o f crabs i n traps that had escape rings and those that had no escape ring's. In the former, an average o f 5.3 ind iv idua ls were k i l l e d per trap i n a 2-month per iod. In the latter, an average o f 17.3 ind iv idua ls were k i l l e d per trap i n a 2-month per iod. A s s u m i n g that mortal i ty rate is un i fo rm throughout the study per iod and throughout the year, annual morta l i ty is and 31.8 and 103.8 crabs per trap respect ively. The average, o f these three values is 53.8 crabs- trap"'year"'. 44 Applied to the previous parameter, the annual mortality from ghost traps is 2,572,693 crabs. The average weight of blue crabs in October 1999 to November 2000 landings was 0.35 lbs, or 0.16 kg (Murphy et al. 2001). The weight of ghost fishing catch is thus 408,441 kg, or 12.1% of the 3,379,000 kg landed in 2001 (A. Jackson, pers. comm. to A. Poon, 2004). Monte Carlo simulation A. McMillen-Jackson (pers. comm. to Guillory et al. 2001) estimates gear loss to be between 30% and 50%, which is 40% ± 25%. The percentage of traps lost to ghost fishing is the grand mean calculated from case studies where the parameters were given, which results in 96.9% + 3.1%/- 51.3% of lost traps (96.9% ± 51.3%, with the upper limit truncated at 100%). The values for the lifespan of ghost fishing traps in this case study was an average of thirteen other case studies, some of which had asymmetrical variances. In the cases where variances were asymmetrical, the average of the variances in each case study was taken and used towards the following formula: Sd((X+Y)/13) = (sqrt(Var(X) + Var(Y)))/13 resulting in a mean of 0.417 + 141%/-100% years (0.417 ± 141%, with the lower limit truncated at 0). The mortality rate is taken as the mean from three values, 53.8 crabs-trap"'year"' ± 34.1%, as explained in the narrative above. The distribution of ghost fishing mortality as a percentage of the reported landed catch is shown in Figure 3.1.17. Gulf of Mexico blue crab fishery 1993 Parameter list Parameter i : Parameter i i : Parameter 1: Parameter 2: Parameter 3: Not provided Not provided 605,000 traps 40% ± 25% 242,000 traps 45 X X X X Parameter 4: Parameter 5: Parameter 6: Parameter 7: Parameter 8: Parameter 9: Parameter 10 Parameter 11: Parameter 12 Resul t : 9 6 . 9 % + 3 .1%/- 51 .3% 234,384 traps 0.417 + 141%/-100% years 97,795 traps 53.8 ± 3 4 . 1 % crabs- trap"'year' 5,261,397 crabs 0.16 k g 835,299 kg 27,722,644 k g 3.0% .-i Narrative T h e N a t i o n a l M a r i n e Fisher ies Service ( N M F S ) estimated that there were 605,000 traps i n use i n 1993 i n F l o r i d a , A l a b a m a , M i s s i s s i p p i , and L o u i s i a n a ( G u i l l o r y et al. 2001) . A . M c M i l l a n - J a c k s o n (pers. c o m m . to G u i l l o r y et al. 2001) estimates the trap loss rate i n F l o r i d a to be 30-50%; for the purposes o f this case study, the mean was used, resul t ing i n an estimate o f 242,000 traps lost annually. The estimate o f trap loss inc luded stolen traps, but there was no ind ica t ion o f the propor t ion o f traps lost to ghost f ishing. Hence , a modi f ie r o f 9 6 . 9 % was appl ied. Th i s modi fe r was the mean calculated f rom other case studies where this parameter was specif ied: 4 6 . 5 % i n the 1987 B C Dungeness crab fishery (Breen 1987); 100% i n the 1990 C h i n i a k B a y red k i n g crab fishery (assumed f rom R . M e y e r , pers c o m m to S m o l o w i t z 1978a); 100% i n the 1976 At l an t i c offshore lobster fishery (assumed f rom S m o l o w i t z 1978b); 100% i n the 1969-1989 G u l f o f St. Lawrence snow crab fishery (assumed f rom V i e n n e a u and M o r i y a s u 1994); 100% i n the 1987 G u l f o f St. L a w r e n c e snow crab fishery ( M a l l e t et al. 1998); 100% i n the 1987 N e w B r u n s w i c k snow crab fishery ( M a l l e t et al. 1998); 100% i n the L o u i s i a n a b lue crab fishery c. 1993 ( G u i l l o r y 1999); 100% i n the 1988 L o u i s i a n a blue crab fishery ( G u i l l o r y 1999); 100% i n the 1993 L o u i s i a n a blue crab fishery ( G u i l l o r y 1999); and 100% i n the 1974 N e w f o u n d l a n d snow crab fishery ( M i l l e r 1977). The result was 234,384 ghost f ish traps. N o estimate for the functional l i fespan o f the traps was available, so an average length o f t ime, 0.417 years, was taken from a l l case studies i n w h i c h ghost f i sh ing pe r iod was specified. A s s u m i n g that the rate o f entry o f lost traps was equal to the rate o f degradation, 97,795 traps are estimated to be ghost f i sh ing inshore i n the G u l f o f M e x i c o annually. In the absence o f morta l i ty rate estimates for the F l o r i d a blue crab fishery, the average o f morta l i ty rates from L o u i s i a n a b lue crab fisheries were appl ied instead. In G u i l l o r y ' s (1993) study o f ghost f i sh ing b y blue crab traps i n L o u i s i a n a , an average o f 25.8 ind iv idua l s died per trap. Arcemen t and G u i l l o r y (1993) performed a compar ison o f 46 morta l i ty o f crabs i n traps that had escape rings and those that had no escape rings. In the former, an average o f 5.3 ind iv idua l s were k i l l e d per trap i n a 2-month per iod . In the latter, an average o f 17.3 ind iv idua l s were k i l l e d per trap i n a 2-month per iod. A s s u m i n g that mortal i ty rate is un i fo rm throughout the study per iod and throughout the year, annual mor ta l i ty is and 31.8 and 103.8 crabs per trap respectively. The average o f these three values is 53.8 crabs- trap^year" 1 . A p p l i e d to the previous parameter, the annual mor ta l i ty f rom ghost traps is 5,261,397 crabs. The average weight o f blue crabs i n October 1999 to N o v e m b e r 2000 landings was 0.35 lbs, or 0.16 k g ( M u r p h y et al. 2001). The weight o f ghost f i sh ing catch is thus 835,299 k g , or 3 .0% o f the combined landings caught i n b lue crab traps i n F l o r i d a , A l a b a m a , M i s s i s s i p p i , and L o u i s i a n a i n 1993 (27,722,644 k g ; N a t i o n a l M a r i n e Fisheries Serv ice 2004). Monte Carlo simulation Gear loss i n this fishery is estimated to be between 3 0 % and 5 0 % ( A . M c M i l l a n -Jackson pers. c o m m . to G u i l l o r y et al. 2001) , or 40%> ± 2 5 % . The percentage o f traps lost to ghost f ishing is the grand mean calculated from case studies where the parameters were g iven , w h i c h results i n 9 6 . 9 % + 3 .1%/- 51 .3% o f lost traps (96.9% ± 51 .3%, w i t h the upper l i m i t truncated at 100%). The values for the l i fespan o f ghost f ishing traps i n this case study was an average o f thirteen other case studies, some o f w h i c h had asymmetr ica l variances. In the cases where variances were asymmetr ica l , the average o f the variances i n each case study was taken and used towards the f o l l o w i n g formula: S d ( ( X + Y ) / 1 3 ) = ( sq r t (Var (X) + Var (Y) ) ) / 13 resul t ing i n a mean o f 0.417 + 141%/-100% years (0.417 ± 1 4 1 % , w i t h the lower l imi t truncated at 0). The mortal i ty rate is taken as the mean f rom three values, 53.8 crabs-trap"'year~1 ± 3 4 . 1 % , as expla ined i n the narrative above. The distr ibution o f ghost f i sh ing morta l i ty as a percentage o f the reported landed catch is shown i n F igure 3.1.18. 47 Gulf of Mexico blue crab fishery 1999 or 2000 Parameter list Parameter i: 5,000 fishers X Parameter ii: 200 traps = Parameter 1: 1,000,000 traps X Parameter 2: 25% = Parameter 3: 250,000 traps X Parameter 4: 96.9%+ 3.1%/-51.3% = Parameter 5: 242,132 traps X Parameter 6: 0.417 + 141%/-100% years = Parameter 7: 101,028 traps X Parameter 8: 53.8 ± 34.1% crabs- trap"'year = Parameter 9: 5,435,327 crabs X Parameter 10: 0.16 kg = Parameter 11: 862,913 kg Parameter 12: 14,757,422 kg = Result: 5.8% Narrative In 1999 or 2000, there were 2,381 licensed commercial blue crab (Callinectes sapidus) fishers in the Florida West Coast, 174 in Alabama, 256 in Mississippi, 3,347 in Louisiana, and 25 in Texas (Guillory et al. 2001). However, since not all license holders actively fish, Guillory et al. (2001) conservatively estimate 5,000 blue crab fishers in the Gulf of Mexico, each fishing an average of 200 traps for a total of 1,000,000 traps. However, the latter number may also be an underestimate, given that Guillory et al. (2001) estimated the following percent frequencies of number of traps per fishers based on a survey of commercial blue crab fishers in the Gulf of Mexico: <200 traps: 200-299 traps 300-399 traps 400-499 traps >500 traps: 33% 29% 17% 9% 12%. 48 G u i l l o r y et al. (2001) then appl ied an annual trap loss rate o f 2 5 % , resul t ing i n an estimate o f 250,000 traps lost annually, but acknowledged that not a l l lost traps continue to fish. A modi f ie r o f 96 .9% was therefore appl ied. T h i s modifer was the mean calculated f rom other case studies where this parameter was specif ied: 4 6 . 5 % i n the 1987 B C Dungeness crab fishery (Breen 1987); 100% i n the 1990 C h i n i a k B a y red k i n g crab fishery (assumed f rom R . M e y e r , pers c o m m to S m o l o w i t z 1978a); 100% i n the 1976 At l an t i c offshore lobster f ishery (assumed f rom S m o l o w i t z 1978b); 100% i n the 1969-1989 G u l f o f St. L a w r e n c e snow crab fishery (assumed f rom V i e n n e a u and M o r i y a s u 1994); 100% i n the 1987 G u l f o f St. Lawrence snow crab fishery (Mal l e t et al. 1998); 100% i n the 1987 N e w B r u n s w i c k snow crab fishery ( M a l l e t et al. 1998); 100% i n the L o u i s i a n a blue crab fishery c. 1993 ( G u i l l o r y 1999); 100% i n the 1988 L o u i s i a n a blue crab fishery ( G u i l l o r y 1999); 100% i n the 1993 L o u i s i a n a b lue crab fishery ( G u i l l o r y 1999); and 100% i n the 1974 N e w f o u n d l a n d snow crab fishery ( M i l l e r 1977). The result was 242,132 ghost f ish traps. N o estimate for the functional l ifespan o f the traps was avai lable, so an average length o f t ime, 0.417 years, was taken from a l l case studies i n w h i c h ghost f i sh ing per iod was specified. The values f rom those case studies were: 2.2 years i n the 1984 B C Dungeness crab fishery (Breen 1987); 0.418 years i n the 1969-70 A l a s k a red k i n g crab fishery ( K i m k e r 1990); 0.375 years ± 1 1 % i n the 1986 C a l i f o r n i a Dungeness crab fishery (Kennedy 1986); 0.245 +44%/-19% years i n the 1992 C o o k Inlet red k i n g crab fishery ( K i m k e r 1990); and 0.243 ± 3 3 % years. The last va lue is a composi te i n itself, obtained from 0.241 +20.8%/-27.1% years ( K i m k e r 1990) and 0.245 +44%/-19% years ( K i m k e r 1990), and used i n 9 case studies: 1993 B e r i n g Sea crab fishery; 1993 B e r i n g Sea red k i n g crab fishery; 1990 B r i s t o l B a y red k i n g crab fishery; 1991 B r i s t o l B a y red k i n g crab fishery; 1994 B r i s t o l B a y red k i n g crab fishery; 1996 B r i s t o l B a y red k i n g crab fishery; C h i n i a k B a y red k i n g crab fishery i n the 1990s; and 2 199.1 eastern B e r i n g Sea snow crab fisheries. A s s u m i n g that the rate o f entry o f lost traps was equal to the rate o f degradation, 101,028 traps were estimated to be ghost f i sh ing i n the G u l f o f M e x i c o annually. In the absence o f morta l i ty rate estimates for the G u l f o f M e x i c o b lue crab fisheries i n general, the average o f morta l i ty rates from L o u i s i a n a blue crab fisheries were appl ied instead. In G u i l l o r y ' s (1993) study o f ghost f i sh ing by blue crab traps i n L o u i s i a n a , an average o f 25.8 ind iv idua ls d ied per trap. Arcement and G u i l l o r y (1993) performed a compar i son o f mortal i ty o f crabs i n traps that had escape rings and those that had no escape r ings. In the former, an average o f 5.3 ind iv idua l s were k i l l e d per trap i n a 2-month per iod. In the latter, an average o f 17.3 ind iv idua l s were k i l l e d per trap i n a 2-month per iod. A s s u m i n g that mortal i ty rate is un i fo rm throughout the study pe r iod and throughout the year, annual mortal i ty is and 31.8 and 103.8 crabs per trap respect ively. The average o f these three values is 53.8 crabs- trap"'year" 1. 49 A p p l i e d to the previous parameter, the annual mortal i ty from ghost traps is 5,435,327 crabs. The average weight o f b lue crabs i n F l o r i d a ' s October 1999 to N o v e m b e r 2000 landings was 0.35 lbs, or 0.16 k g ( M u r p h y et al. 2001) . T h e weight o f ghost f i sh ing catch is thus 862,913 k g , or 5.8% o f the combined landings o f b lue crabs (exc lud ing soft crabs and peelers) caught i n blue crab traps i n F l o r i d a , A l a b a m a , M i s s i s s i p p i , and L o u i s i a n a i n 2001 (14,757,422 k g ; Na t iona l M a r i n e Fisher ies Service 2004). Monte Carlo simulation T h e percentage o f traps lost to ghost f i sh ing is the grand mean calculated f rom case studies where the parameters were g iven , w h i c h results i n 9 6 . 9 % + 3 .1%/- 51 .3% o f lost traps (96 .9% ± 51 .3%, w i t h the upper l i m i t truncated at 100%). The values for the l ifespan o f ghost f i sh ing traps i n this case study was an average o f thirteen other case studies, some o f w h i c h had asymmetr ica l variances. In the cases where variances were asymmetr ica l , the average o f the variances i n each case study was taken and used towards the f o l l o w i n g formula : S d ( ( X + Y ) / 1 3 ) = (sqr t (Var (X) + V a r ( Y ) ) ) / 1 3 resul t ing i n a mean o f 0.417 + 141%/-100% years (0.417 ± 141%, w i t h the lower l imi t truncated at 0). T h e morta l i ty rate is taken as the mean f rom three values, 53.8 crabs-trap~1year*1 ± 3 4 . 1 % , as expla ined i n the narrative above. The dis t r ibut ion o f ghost f i sh ing mor ta l i ty as a percentage o f the reported landed catch is shown i n F igure 3.1.19. Gulf of St. Lawrence snow crab fishery 1980 Parameter list Parameter i : 130 fishers Parameter i i : 150 Parameter 1: 19,500 traps Parameter 2: 10.5% ± 9 0 % Parameter 3: 2,048 traps Parameter 4: 96 .9% + 3 .1%/- 5 1 . 3 % Parameter 5: 1,983 traps 50 0 .417+ 141 % / - 1 0 0 % years 827 traps 41 crabs- trap"'year" 1 Parameter 6 Parameter 7 Parameter 8 Parameter 9: 33,924 crabs Parameter 10: 0.572 k g Parameter 11: 19,405 kg Parameter 12: 11,782,000 k g Resul t : 0.16% Narrative In 1980, there were 130 l icensed fishers i n the southwestern G u l f o f St. Lawrence snow crab (Chionoecetes opilio) f ishery ( B l o i s 1992). There is no estimate o f h o w many traps they used on average, but each fisher is a l l owed a m a x i m u m o f 150 traps ( M . M o r i y a s u , pers. c o m m . to A . P o o n , 2004) . A s s u m i n g a l l fishers used the m a x i m u m number o f traps, 19,500 traps were used. N o estimate for annual rate o f loss was p rov ided , so the average f rom case studies (20% i n W . N i p p e s pers. c o m m . to K r u s e and K i m k e r 1993 and 1% i n R . M o r r i s o n pers. c o m m . to Stevens et al. 2000) was used for the snow crab fisheries i n A l a s k a was appl ied, resul t ing i n 2,048 traps lost per year. N o t a l l lost traps remain capable o f ghost f i shing, so a modi f ie r o f 9 6 . 9 % was therefore appl ied. Th i s modifer was the mean calculated from other case studies where this parameter was specif ied: 4 6 . 5 % i n the 1987 B C Dungeness crab fishery (Breen 1987); 100% i n the 1990 C h i n i a k B a y red k i n g crab fishery (assumed f rom R . M e y e r , pers c o m m to S m o l o w i t z 1978a); 100% i n the 1976 At l an t i c offshore lobster fishery (assumed f rom S m o l o w i t z 1978b); 100% i n the 1969-1989 G u l f o f St. L a w r e n c e snow crab fishery (assumed f rom V i e n n e a u and M o r i y a s u 1994); 100% i n the 1987 G u l f o f St. L a w r e n c e snow crab fishery (Ma l l e t et al. 1998); 100% i n the 1987 N e w B r u n s w i c k snow crab fishery (Mal l e t et al. 1998); 100% i n the L o u i s i a n a blue crab fishery c. 1993 ( G u i l l o r y 1999); 100% i n the 1988 L o u i s i a n a blue crab fishery ( G u i l l o r y 1999); 100% i n the 1993 L o u i s i a n a blue crab fishery ( G u i l l o r y 1999); and 100% i n the 1974 N e w f o u n d l a n d snow crab fishery ( M i l l e r 1977). The result was 1,983 ghost f ish traps. N o estimate for the functional l i fespan o f the traps was avai lable , so an average length o f t ime, 0.417 years, was taken f rom a l l case studies i n w h i c h ghost f i sh ing per iod was specif ied. The values from those case studies were: 2.2 years i n the 1984 B C Dungeness crab fishery (Breen 1987); 0.418 years i n the 1969-70 A l a s k a red k i n g crab fishery ( K i m k e r 1990); 0.375 years ± 1 1 % i n the 1986 C a l i f o r n i a Dungeness crab fishery ( K e n n e d y 1986); 0.245 +44%/-19% years i n the 1992 C o o k Inlet red k i n g crab fishery ( K i m k e r 1990); and 0.243 ± 3 3 % years. The last value is a composi te i n itself, obtained f rom 0.241 +20.8%/-27.1% years ( K i m k e r 1990) and 0.245 +44%/-19% years ( K i m k e r 51 1990), and used i n 9 case studies: 1993 B e r i n g Sea crab fishery; 1993 B e r i n g Sea red k i n g crab fishery; 1990 B r i s t o l B a y red k i n g crab fishery; 1991 B r i s t o l B a y red k i n g crab fishery; 1994 B r i s t o l B a y red k i n g crab fishery; 1996 B r i s t o l B a y red k i n g crab fishery; C h i n i a k B a y red k i n g crab fishery i n the 1990s; and 2 1991 eastern B e r i n g Sea snow crab fisheries. A s s u m i n g that the rate o f entry o f lost traps was equal to the rate o f degradation, 827 traps are estimated to be ghost f i sh ing i n the G u l f o f St. L a w r e n c e annually. V i e n n e a u and M o r i y a s u (1994) determined that the mor ta l i ty rate o f crab caught b y a lost trap is 5 0 % that o f regular ly funct ioning traps; appl ied to the observed C P U E i n the Sp r ing 1993 fishery o f 82 ind iv idua l s per trap, each trap w o u l d k i l l 41 crabs per year. The 827 lost traps thus w o u l d k i l l an average o f 33,924 snow crabs each year. V i e n n e a u and M o r i y a s u (1994), based on unpubl ished data from Y . Ch ias son , D F O G u l f R e g i o n , state that the average i nd iv idua l weight is 0.572 k g . The weight o f ghost f ishing catch is thus 19,405 k g , 0 .16% o f the reported 11,782,000 k g landed i n 1987 ( M a l l e t et al. 1998). Monte Carlo simulation A n estimate o f gear loss was not avai lable for this fishery, so the average f rom other case studies, both i n the B e r i n g Sea, was taken. P r e l i m i n a r y informat ion from R . M o r r i s o n (pers. c o m m . to Stevens et al. 2000) estimate the gear loss rate at 1%, w h i l e P a u l et al. (1994) estimated the gear loss rate at 2 0 % , w h i c h results i n a midpo in t estimate o f 10 .5% ± 90%. The percentage o f traps lost to ghost f ishing is the grand mean calculated f rom case studies where the parameters were g iven , w h i c h results i n 9 6 . 9 % + 3 .1%/- 51 .3% o f lost traps (96.9% ± 51 .3%, w i t h the upper l imi t truncated at 100%). The values for the l i fespan o f ghost f i sh ing traps i n this case study was an average o f thirteen other case studies, some o f w h i c h had asymmetr ica l variances. In the cases where variances were asymmetr ica l , the average o f the variances i n each case study was taken and used towards the f o l l o w i n g formula: S d ( ( X + Y ) / 1 3 ) = ( sq r t (Var (X) + V a r ( Y ) ) ) / 1 3 resul t ing i n a mean o f 0.417 + 141%/-100% years (0.417 ± 1 4 1 % , w i t h the lower l imi t truncated at 0). The dis tr ibut ion o f ghost f i sh ing morta l i ty as a percentage o f the reported landed catch is shown i n F igure 3.1.20. 52 Louisiana blue crab fishery 1999 or 2000 Parameter list Parameter i : 3,347 fishers X Parameter i i : 257 traps Parameter 1: Not applicable Parameter 2: Not applicable = Parameter 3: 860,179 traps X Parameter 4: 96.9%+ 3.1%/-51.3% Parameter 5: 833,109 traps - Parameter 6: 0.417 + 141%/-100% years = Parameter 7: 347,610 traps X Parameter 8: 53.8 ± 34.1% crabs- trap"'year' = Parameter 9: 18,701,417 crabs X Parameter 10: 0.16 kg = Parameter 11: 2,969,037 kg -r- Parameter 12: 18,897,136 kg = Result: 15.7% Narrative In 1999 or 2000, there were 3,347 licensed commercial blue crab (Callinectes sapidus) fishers, and 257 traps per fisher were lost or stolen (Guillory and Merrell 1998 cited in Guillory et al. 2001), resulting in 860,179 lost traps per year. There is no estimate for the proportion of those traps that are actually ghost fishing, so a modifier of 96.9% was applied. This modifer was the mean calculated from other case studies where this parameter was specified: 46.5% in the 1987 BC Dungeness crab fishery (Breen 1987); 100% in the 1990 Chiniak Bay red king crab fishery (assumed from R. Meyer, pers comm to Smolowitz 1978a); 100% in the 1976 Atlantic offshore lobster fishery (assumed from Smolowitz 1978b); 100% in the 1969-1989 Gulf of St. Lawrence snow crab fishery (assumed from Vienneau and Moriyasu 1994); 100% in the 1987 Gulf of St. Lawrence snow crab fishery (Mallet et al. 1998); 100% in the 1987 New Brunswick snow crab fishery (Mallet et al. 1998); 100% in the Louisiana blue crab fishery c. 1993 (Guillory 1999); 100% in the 1988 Louisiana blue crab fishery (Guillory 1999); 100% in the 1993 Louisiana blue crab fishery (Guillory 1999); and 100% in the 1974 Newfoundland snow crab fishery (Miller 1977). Thus, 833,109 ghost traps are added to Louisiana's blue crab fishing grounds each year. 53 In the absence o f an estimate for the funct ional l i fe o f the ghost traps, an average o f ghost f ishing pe r iod (0.417 years) was taken f rom a l l other case studies that had a value for this parameter was used. The values f rom those case studies were: 2.2 years i n the 1984 B C Dungeness crab fishery (Breen 1987); 0.418 years i n the 1969-70 A l a s k a red k i n g crab fishery ( K i m k e r 1990); 0.375 years ± 1 1 % i n the 1986 C a l i f o r n i a Dungeness crab fishery (Kennedy 1986); 0.245 +44%/-19% years i n the 1992 C o o k Inlet red k i n g crab fishery ( K i m k e r 1990); and 0.243 ± 3 3 % years. The last value is a composi te i n itself, obtained from 0.241 +20.8%/-27 .1% years ( K i m k e r 1990) and 0.245 +44%/-19% years ( K i m k e r 1990), and used i n 9 case studies: 1993 B e r i n g Sea crab fishery; 1993 B e r i n g Sea red k i n g crab fishery; 1990 B r i s t o l B a y red k i n g crab fishery; 1991 B r i s t o l B a y red k i n g crab fishery; 1994 B r i s t o l B a y red k i n g crab fishery; 1996 B r i s t o l B a y red k i n g crab fishery; C h i n i a k B a y red k i n g crab fishery i n the 1990s; and 2 1991 eastern B e r i n g Sea snow crab fisheries. A s s u m i n g that the rate o f entry o f lost traps was equal to the rate o f degradation, 347,610 traps are estimated to be ghost f i sh ing i n the area annually. In G u i l l o r y ' s (1993) study o f ghost f i sh ing b y blue crab traps i n L o u i s i a n a , an average o f 25.8 ind iv idua l s died per trap. A r c e m e n t and G u i l l o r y (1993) performed a compar i son o f mor ta l i ty o f crabs i n traps that had escape rings and those that had no escape rings. In the former, an average o f 5.3 ind iv idua l s were k i l l e d per trap i n a 2-month per iod. In the latter, an average o f 17.3 ind iv idua l s were k i l l e d per trap i n a 2-month per iod. A s s u m i n g that mortal i ty rate is un i fo rm throughout the study pe r iod and throughout the year, annual mortal i ty is and 31.8 and 103.8 crabs per trap respect ively. The average o f these three values is 53.8 crabs- trap"'year"'. A p p l i e d to the previous parameter, the annual morta l i ty f rom ghost traps is 18,701,417 crabs. The average weight o f b lue crabs i n October 1999 to N o v e m b e r 2000 landings was 0.35 lbs, or 0.16 k g ( M u r p h y et al. 2001). The weight o f ghost f i sh ing catch is thus 2,969,037 k g , or 15.7% o f the reported 18,897,136 k g landed i n 2001 (Na t iona l M a r i n e Fisher ies Service 2004). Monte Carlo simulation The percentage o f traps lost to ghost f i sh ing is the grand mean calculated from case studies where the parameters were g iven , w h i c h results i n 96 .9% + 3 .1%/- 5 1 . 3 % o f lost traps (96 .9% ± 51 .3%, wi th the upper l i m i t truncated at 100%). The values for the lifespan o f ghost f i sh ing traps i n this case study was an average o f thirteen other case studies, some o f w h i c h had asymmetr ica l variances. In the cases where variances were asymmetr ica l , the average o f the variances i n each case study was taken and used towards the f o l l o w i n g formula : 54 Sd((X+Y)/13) - (sqrt(Var(X) + Var(Y)))/13 resulting in a mean of 0.417 + 141%/-100% years (0.417 ± 141%, with the lower limit truncated at 0). The mortality rate is taken as the mean from three values, 53.8 crabs-trap'year"1 ± 34.1%, as explained in the narrative above. The distribution of ghost fishing mortality as a percentage of the reported landed catch is shown in Figure 3.1.21. Louisiana blue crab fishery 1993 Parameter list Parameter i : 2,783 ± 3% fishers X Parameter i i : 164 traps = Parameter 1: 453,548 traps X Parameter 2: 10% trap = Parameter 3: 45,355 traps X Parameter 4: 100% = Parameter 5: 45,355 traps X Parameter 6: 0.417 + 141%/-100% years = Parameter 7: 18,924 traps X Parameter 8: 53.8 ± 34.1% crabs- trap"'year = Parameter 9: 1,108,112 crabs X Parameter 10: 0.16 kg = Parameter 11: 161,635kg -7- Parameter 12: 20,796,199 kg = Result: 0.78% . Narrative In 1993, 2,854 licenses were issued in the Louisiana blue crab (Callinectes sapidus) fishery; 2,711 of those were estimated to be in full-time use (Guillory et al. 1996). For the purposes of this case study, the mean of the two numbers (2,783) was used to reflect both the full-time and part-time fishers. A n average of 163 traps was used by each fisher, taking the total number of traps used that year to 453,548. Guil lory (1999) estimates an annual trap loss rate of 10% in the fishery; 45,355 traps would be added to the ghost fishing fleet each year. Guillory (1999) assumes that all these lost traps are 55 ghost f i shing, but gives no estimate for the duration o f their functional l ives , so the average (0.417 years) was taken f rom other case studies where this parameter was specif ied and appl ied. The values from those case studies were: 2.2 years in the 1984 B C Dungeness crab fishery (Breen 1987); 0:418 years i n the 1969-70 A l a s k a red k i n g crab fishery ( K i m k e r 1990); 0.375 years ± 1 1 % i n the 1986 C a l i f o r n i a Dungeness crab fishery ( K e n n e d y 1986); 0.245 +44%/-19% years i n the 1992 C o o k Inlet red k i n g crab fishery ( K i m k e r 1990); and 0.243 ± 3 3 % years. The last value is a composi te i n itself, obtained f rom 0.241 +20.8%/-27.1% years ( K i m k e r 1990) and 0.245 +44%/-19% years ( K i m k e r 1990), and used i n 9 case studies: 1993 B e r i n g Sea crab fishery; 1993 B e r i n g Sea red k i n g crab fishery; 1990 B r i s t o l B a y red k i n g crab fishery; 1991 B r i s t o l B a y red k i n g crab fishery; 1994 B r i s t o l B a y red k i n g crab fishery; 1996 B r i s t o l B a y red k i n g crab fishery; C h i n i a k B a y red k i n g crab fishery i n the 1990s; and 2 1991 eastern B e r i n g Sea snow crab fisheries. A s s u m i n g that the rate o f entry o f lost traps was equal to the rate o f degradation, 18,924 traps are estimated to be ghost f i sh ing i n the area annually. In G u i l l o r y ' s (1993) study o f ghost f i sh ing b y blue crab traps i n Lou i s i ana , an average o f 25.8 ind iv idua l s d ied per trap. Arcemen t and G u i l l o r y (1993) performed a compar i son o f mor ta l i ty o f crabs i n traps that had escape r ings and those that had no escape rings. l a t h e former, an average o f 5.3 ind iv idua l s were k i l l e d per trap i n a 2-month per iod. l a t h e latter, an average o f 17.3 ind iv idua l s were k i l l e d per trap i n a 2-month per iod. A s s u m i n g that mor ta l i ty rate is un i fo rm throughout the study per iod and throughout the year, annual mor ta l i ty is and 31.8 and 103.8 crabs per trap respectively. The average o f these three values is 53.8 crabs- trap _ 1year"' . A p p l i e d to the previous parameter, the annual mor ta l i ty f rom ghost traps 1,018,112 crabs. The average weight o f b lue crabs i n October 1999 to N o v e m b e r 2000 landings was 0.35 lbs, or 0.16 k g ( M u r p h y et al. 2001) . The weight o f ghost f i sh ing catch is thus 161,635 k g , or 0 .78% o f the reported 20,796,199 k g landed i n 1993 ( N M F S ci ted i n G u i l l o r y et al. 1996). Monte Carlo simulation Since 2854 licenses were issued i n the L o u i s i a n a blue crab fishery i n 1993 but o n l y 2711 trappers were w o r k i n g fu l l t ime ( G u i l l o r y et al. 1996), the midpoin t (2783) was taken to represent the part-time fishers, w i t h the error ( ± 3%) representing the lower and upper l imi ts o f the value. The values for the l i fespan o f ghost f ishing traps in this case study was an average o f thirteen other case studies, some o f w h i c h had asymmetr ica l variances. In the cases where variances were asymmetr ica l , the average o f the variances in each case study was taken and used towards the f o l l o w i n g formula: 56 S d ( ( X + Y ) / 1 3 ) = ( sqr t (Var (X) + V a r ( Y ) ) ) / 1 3 resul t ing i n a mean o f 0.417 + 141%/-100% years (0.417 ± 1 4 1 % , w i t h the lower l imi t truncated at 0). The morta l i ty rate is taken as the mean f rom three values, 53.8 crabs-trap" 1year" 1 ± 3 4 . 1 % , as expla ined i n the narrative above. The dis t r ibut ion o f ghost f ishing morta l i ty as a percentage o f the reported landed catch is shown i n F igu re 3.1.22. Ireland spider crab fishery 2000 Parameter list Parameter i : N o t p rov ided Parameter i i : N o t p rov ided Parameter 1: 10,000 traps X Parameter 2: 18.6% ± 6 . 5 % = Parameter 3: 1,862 traps X Parameter 4: 9 6 . 9 % I 3 .1%/- 51 .3% = Parameter 5: 1,804 traps X Parameter 6: 0.417 + 141%/-100% years = Parameter 7: 753 traps X Parameter 8: 7.08 crabs- trap"'year _ 1 = Parameter 9: 5,329 crabs X Parameter 10: 0.70 k g = Parameter 11: 3,370 kg Parameter 12: 402,700 k g - Resul t : 0.93% Narrative In 2000, an estimated 10,000 traps fished for spider crabs (Maja squinado) i n Tralee and B r a n d o n B a y s i n Ireland (Fahy 2001). N o estimate for the rate o f loss for traps was avai lable, and since no other case studies i n this paper are s imi lar enough to assume a s imi la r rate o f loss, the average annual loss rate (18.6%) from a l l case studies w i t h this parameter was used instead, result ing i n 1,862 lost traps per year. The case studies were: 3 5 % ± 4 3 % i n the 1999/2000 A l a b a m a blue crab fishery ( L . Har tman pers. c o m m to G u i l l o r y et al. 2001); 2 0 % used i n the 1993 B e r i n g Sea crab fishery ( W . N i p p e s 57 to K r u s e and K i m k e r 1993); 5% ± 3 3 % i n the 1993 B e r i n g Sea red k i n g crab fishery ( W . N i p p e s pers. c o m m . to Stevens et al. 1993); 10% i n the 1996 B r i s t o l B a y red k i n g crab fishery (Stevens 1996); 1% trap loss rate used i n the 1999 B e r i n g Sea snow crab fishery (R. M o r r i s o n pers. c o m m . to Stevens et al. 2000) ; 7 .5% ± 3 3 % i n the 1987 A t l a n t i c inshore lobster f ishery (Krouse 1989 ci ted i n B r e e n 1990); 2 5 % ± 2 0 % i n the 1976 At l an t i c offshore lobster fishery ( S m o l o w i t z 1979b); 3 3 % i n the 1973 At l an t i c seacoast inshore lobster f ishery (Prudden 1962); 10 .9% i n the 1984 B C Dungeness crab fishery (Breen 1987); 17.6% i n the 1975/1976 C o l u m b i a R i v e r estuary Dungeness crab fishery ( M u i r et al. 1984); 4 0 % ± 2 5 % i n the 2001 F l o r i d a b lue crab fishery ( A . M c M i l l a n -Jackson pers. c o m m . to G u i l l o r y et al. 2001); 2 5 % i n the 2001 G u l f o f M e x i c o blue crab fishery ( G u i l l o r y et al. 2001) ; 10% i n the 1999 L o u i s i a n a blue crab fishery ( G u i l l o r y 1999); 2 2 . 5 % ± 1 1 % i n the 1975 M a i n e lobster f ishery (Sheldon and D o w 1975); 7 .5% ± 3 3 % used i n the 1992 M a i n e lobster fishery calculat ions (Carr and Har r i s 1987); 25% ± 2 0 % (T. F l o y d pers. c o m m . to G u i l l o r y et al. 2001) ; 7 .5% ± 3 3 % i n the 1987 N e w E n g l a n d lobster f ishery (Breen 1990); 8.3% i n the 1974 N e w f o u n d l a n d snow crab fishery (calculated f rom 1,000 traps lost f rom 12,000 vessels; M i l l e r 1977); 10% i n the 1971/1972 Pac i f i c Coast Dungeness crab fishery (Pac i f ic M a r i n e Fisheries C o m m i s s i o n 1978); 12 .5% ± 2 0 % used i n the 1980 R h o d e Is land Lobs te r fishery (Fogarty and B o r d e n 1980 ci ted i n B r e e n 1990) 4 2 . 5 % ± 18% i n the 1999 or 2000 Texas blue crab fishery (Texas Parks and W i l d l i f e Department unpubl ished data c i ted i n G u i l l o r y et al. 2001) ; 11.9% i n the Wash ing ton Dungeness crab fishery (Bar ry pers. c o m m . to B r e e n 1990); 15.7% i n the 1972 W a s h i n g t o n Dungeness crab fishery (Tegelberg 1974); 15 .7% i n the 1973 Wash ing ton Dungeness crab fishery (Tegelberg 1974). N o t a l l lost traps are capable o f ghost f i sh ing; some are lost to vanda l i sm or are stolen, so a modi f ie r o f 9 6 . 9 % was applied. T h i s modi fe r was the mean calculated f rom other case studies where this parameter was specif ied: 4 6 . 5 % i n the 1987 B C Dungeness crab fishery (Breen 1987); 100% i n the 1990 C h i n i a k B a y red k i n g crab fishery (assumed from R . M e y e r , pers c o m m to S m o l o w i t z 1978a); 100% i n the 1976 At l an t i c offshore lobster fishery (assumed f rom S m o l o w i t z 1978b); 100% i n the 1969-1989 G u l f o f St. Lawrence snow crab fishery (assumed f rom V i e n n e a u and M o r i y a s u 1994); 1 0 0 % i n the 1987 G u l f o f St. L a w r e n c e snow crab fishery ( M a l l e t et al. 1998); 100% i n the 1987 N e w B r u n s w i c k snow crab fishery (Mal l e t et al. 1998); 100% i n the L o u i s i a n a blue crab fishery c. 1993 ( G u i l l o r y 1999); 100% i n the 1988 L o u i s i a n a blue crab fishery ( G u i l l o r y 1999); 100% i n the 1993 L o u i s i a n a blue crab fishery ( G u i l l o r y 1999); and 100% i n the 1974 N e w f o u n d l a n d snow crab fishery ( M i l l e r 1977). T h i s results i n 1,804 lost traps capable o f ghost f ishing. There was no ind ica t ion o f h o w long the lost traps w o u l d remain funct ional , so an average length o f t ime, 0.417 years, was taken f rom a l l case studies i n w h i c h ghost f ishing per iod was specif ied. The values from those case studies were: 2.2 years i n the 58 1984 B C Dungeness crab fishery (Breen 1987); 0.418 years i n the 1969-70 A l a s k a red k i n g crab fishery ( K i m k e r 1990); 0.375 years ± 1 1 % i n the 1986 C a l i f o r n i a Dungeness crab fishery (Kennedy 1986); 0.245 +44%/-19% years in the 1992 C o o k Inlet red k i n g crab fishery ( K i m k e r 1990); and 0.243 ± 3 3 % years. The last value is a composi te i n itself, obtained f rom 0.241 +20.8%/-27.1% years ( K i m k e r 1990) and 0.245 +44%/-19% years ( K i m k e r 1990), and used in 9 case studies: 1993 B e r i n g Sea crab fishery; 1993 B e r i n g Sea red k i n g crab fishery; 1990 B r i s t o l B a y red k i n g crab fishery; 1991 B r i s t o l B a y red k i n g crab fishery; 1994 B r i s t o l B a y red k i n g crab fishery; 1996 B r i s t o l B a y red k i n g crab fishery; C h i n i a k B a y red k i n g crab fishery i n the 1990s; and 2 1991 eastern B e r i n g Sea snow crab fisheries. A s s u m i n g that the rate o f entry o f lost traps was equal to the rate o f degradation, and that the numbers remained constant since then, 753 traps are estimated to be ghost f ishing inshore i n the area annual ly . In a study o f catches f rom a fleet o f 'ghost f i sh ing ' traps where traps were deliberately sunk and then retrieved o f f the coast o f Wales , an average o f 7.08 spider crabs were caught per trap per year ( B u l l i m o r e et al. 2001). Ghost f i sh ing mor ta l i ty from a l l lost traps is thus 5,329 ind iv idua ls per year. W i t h the average weight o f spider crabs at 0.70 k g (Industry Canada 2004), the weight o f ghost catches is 3,730 k g , 0 .93% o f the reported 402,700 k g landed i n Tralee and B r a n d o n B a y s i n 1999 (Fahy 2001) . Monte Carlo simulation T h e value for the trap loss rate i n this case study was the mean o f a l l other case studies w i t h this parameter. T o f ind the var iance o f the mean, the f o l l o w i n g fo rmula was used: S d ( ( X + Y ) / 2 3 ) = ( sqr t (Var (X) + V a r ( Y ) ) ) / 2 3 result ing i n 18.6% ± 6.5%. The variance o f the mean appears to be smal ler than the i n d i v i d u a l l y l is ted variances because there were numerous case studies for w h i c h variances were not g iven and assumed to be zero. The value for the l ifespan o f ghost f i sh ing traps i n this case study was an average o f thirteen other case studies, some o f w h i c h had asymmetr ica l variances. In the cases where variances were asymmetr ica l , the average o f the variances in each case study was taken and used towards the f o l l o w i n g formula : S d ( ( X + Y ) / 1 3 ) = ( sqr t (Var (X) + V a r ( Y ) ) ) / f 3 59 resul t ing i n a mean o f 0.417 + 141%/-100% years (0.417 ± 141%, w i t h the lower l imi t truncated at 0). The dis tr ibut ion o f ghost f ishing mortal i ty as a percentage o f the reported landed catch is shown i n F igure 3.1.23. Ireland brown crab fishery 1968-1972 Parameter list Parameter i : 28.5 ± 9.6% vessels X Parameter i i : 106 ± 11.6% traps Parameter 1: 3,021 traps / Parameter 2: 18.6 ± 6 . 5 % = Parameter 3: 563 traps X Parameter 4: 9 6 . 9 % + 3 .1%/- 51 .3% Parameter 5: 545 traps / Parameter 6: 0.417 + 141%/-100% years = Parameter 7: 227 traps X Parameter 8: 6.06 crabs- t r a p " ' y e a r ' = Parameter 9: 1,378 crabs X Parameter 10: 0.80 k g = Parameter 11: 1,102 kg -f- Parameter 12: 7,774,000 k g = Resul t : 0.01% Narrative In the per iod between 1968-1972, 26-31 vessels each fished 95-117 b r o w n crab (Cancer pagurus) traps i n southeast Ireland ( G i b s o n 1973 ci ted i n F a h y et al. 2002). I f the midranges o f those values were taken, a total o f 3,021 traps were used per year. N o estimate for the rate o f loss for traps was avai lable, and since no other case studies i n this paper are s imi la r enough to assume a s imi la r rate o f loss, the average annual loss rate (18.6%) f rom a l l case studies w i t h this parameter was used instead, result ing i n 563 lost traps per year. The case studies were: 3 5 % ± 4 3 % i n the 1999/2000 A l a b a m a blue crab fishery ( L . Ha r tman pers. c o m m to G u i l l o r y et al. 2001); 2 0 % used i n the 1993 B e r i n g Sea crab fishery ( W . N i p p e s to K r u s e and K i m k e r 1993); 5% ± 3 3 % i n the 1993 B e r i n g Sea red k i n g crab fishery ( W . N i p p e s pers. c o m m . to Stevens et al. 1993); 10% i n the 1996 B r i s t o l B a y red k i n g crab fishery (Stevens 1996); 1% trap loss rate used i n the 1999 B e r i n g Sea snow crab fishery (R. M o r r i s o n pers. c o m m . to Stevens et al. 2000); 7 .5% ± 60 3 3 % i n the 1987 At l an t i c inshore lobster fishery (Krouse 1989 ci ted i n Breen 1990); 2 5 % ± 2 0 % i n the 1976 At l an t i c offshore lobster fishery ( S m o l o w i t z 1979b); 3 3 % i n the 1973 At l an t i c seacoast inshore lobster fishery (Prudden 1962); 10 .9% i n the 1984 B C Dungeness crab fishery (Breen 1987); 17.6% i n the 1975/1976 C o l u m b i a R i v e r estuary Dungeness crab fishery ( M u i r et al. 1984); 4 0 % ± 2 5 % i n the 2001 F l o r i d a blue crab fishery ( A . M c M i l l a n - J a c k s o n pers. c o m m . to G u i l l o r y et al. 2001) ; 2 5 % i n the 2001 G u l f o f M e x i c o blue crab fishery ( G u i l l o r y et al. 2001); 10% i n the 1999 L o u i s i a n a blue crab fishery ( G u i l l o r y 1999); 2 2 . 5 % ± 1 1 % i n the 1975 M a i n e lobster f ishery (Sheldon and D o w 1975); 7 .5% ± 3 3 % used i n the 1992 M a i n e lobster f ishery calculat ions (Carr and Har r i s 1987); 2 5 % ± 2 0 % (T. F l o y d pers. c o m m . to G u i l l o r y et al. 2001); 7 .5% ± 3 3 % i n the 1987 N e w E n g l a n d lobster fishery (Breen 1990); 8 .3% i n the 1974 Newfound land snow crab fishery (calculated f rom 1,000 traps lost f rom 12,000 vessels; M i l l e r 1977); 10% i n the 1971/1972 Pac i f i c Coast Dungeness crab fishery (Pac i f ic M a r i n e Fisheries C o m m i s s i o n 1978); 12 .5% ± 2 0 % used i n the 1980 R h o d e Is land Lobs te r fishery (Fogarty and B o r d e n 1980 c i ted i n B r e e n 1990) 4 2 . 5 % ± 18% i n the 1999 or 2000 Texas blue crab fishery (Texas Parks and W i l d l i f e Department unpubl i shed data c i ted i n G u i l l o r y et al. 2001); 11 .9% i n the Wash ing ton Dungeness crab fishery (Bar ry pers. c o m m . to B r e e n 1990); 15.7% i n the 1972 Wash ing ton Dungeness crab fishery (Tegelberg 1974); 15.7% i n the 1973 Wash ing ton Dungeness crab fishery (Tegelberg 1974) N o t a l l lost traps are capable o f ghost f ishing; some are lost to vanda l i sm or are stolen, so a modi f ie r o f 9 6 . 9 % was appl ied. Th i s modifer was the mean calculated f rom other case studies where this parameter was specified: 4 6 . 5 % i n the 1987 B C Dungeness crab fishery (Breen 1987); 100% i n the 1990 C h i n i a k B a y red k i n g crab fishery (assumed f rom R . M e y e r , pers c o m m to S m o l o w i t z 1978a); 100% i n the 1976 At l an t i c offshore lobster fishery (assumed f rom S m o l o w i t z 1978b); 100% i n the 1969-1989 G u l f o f St. L a w r e n c e snow crab fishery (assumed f rom V i e n n e a u and M o r i y a s u 1994); 100% i n the 1987 G u l f o f St. Lawrence snow crab fishery (Ma l l e t et al. 1998); 100% i n the 1987 N e w B r u n s w i c k snow crab fishery ( M a l l e t et al. 1998); 100% i n the L o u i s i a n a blue crab fishery c. 1993 ( G u i l l o r y 1999); 100% i n the 1988 L o u i s i a n a blue crab fishery ( G u i l l o r y 1999); 100% i n the 1993 L o u i s i a n a blue crab fishery ( G u i l l o r y 1999); and 100% i n the 1974 N e w f o u n d l a n d snow crab fishery ( M i l l e r 1977). T h i s results i n 545 lost traps capable o f ghost f ishing. There was no ind ica t ion o f h o w long the lost traps w o u l d remain functional, so an average length o f t ime, 0.417 years, was taken from a l l case studies i n w h i c h ghost f i sh ing per iod was specified. T h e values from those case studies were: 2.2 years i n the 1984 B C Dungeness crab fishery (Breen 1987); 0.418 years i n the 1969-70 A l a s k a red k i n g crab fishery ( K i m k e r 1990); 0.375 years ± 1 1 % i n the 1986 C a l i f o r n i a Dungeness crab fishery (Kennedy 1986); 0.245 +44%/-19% years i n the 1992 C o o k Inlet red k i n g 61 crab fishery ( K i m k e r 1990); and 0.243 ± 3 3 % years. T h e last value is a composi te i n itself, obtained from 0.241 +20.8%/-27.1% years ( K i m k e r 1990) and 0.245 +44%/-19% years ( K i m k e r 1990), and used i n 9 case studies: 1993 B e r i n g Sea crab fishery; 1993 B e r i n g Sea red k i n g crab fishery; 1990 B r i s t o l B a y red k i n g crab fishery; 1991 B r i s t o l B a y red k i n g crab fishery; 1994 B r i s t o l B a y red k i n g crab fishery; 1996 B r i s t o l B a y red k i n g crab fishery; C h i n i a k B a y red k i n g crab fishery i n the 1990s; and 2 1991 eastern B e r i n g Sea snow crab fisheries. A s s u m i n g that the rate Of entry o f lost traps was equal to the rate o f degradation, and that the numbers remained constant since then, 227 traps are estimated to be ghost f i sh ing inshore in the area annual ly . In a study o f catches f rom a fleet o f 'ghost f i sh ing ' traps where traps were deliberately sunk and then retr ieved o f f the coast o f W a l e s , an average o f 6.06 b r o w n crabs were caught per trap per year ( B u l l i m o r e et al. 2001) . Ghos t f ishing morta l i ty f rom a l l lost traps is thus 1,378 ind iv idua l s per year. W i t h the average weight o f b r o w n crabs at 0.80 k g (R. D e Forges , pers. c o m m . to A . P o o n , 2004) , the weight o f ghost catches is 1,102 k g , 0 .01% o f the reported 7,774,000 k g landed i n southeast Ireland i n 1999. B u l l i m o r e ' s traps caught 0.44 lobsters per trap per year as bycatch; appl ied to the southeast Ir ish fleet, this results i n 93 ind iv idua l s or 55 k g (wi th the average weight o f lobsters at 0.585 k g ( M e r c e r 2002) per year; 0 .44% o f the reported 13,412 k g landed per year i n N o r f o l k i n the late f rom 1969-1973 ( B r o w n 1979). Monte Carlo simulation Tota l gear was calculated f rom the midranges o f 26-31 vessels and 95-117 traps per vessel . The result was 3,021 traps ± 19%, where in the variance was calculated w i t h the f o l l o w i n g formula : S d ( X * Y ) = ( X * Y ) ( S d X / X + S d Y / Y ) The value for the trap loss rate i n this case study was the mean o f a l l other case studies w i t h this parameter. T o f ind the variance o f the mean, the f o l l o w i n g fo rmula was used: S d ( ( X + Y ) / 2 3 ) = ( sq r t (Var (X) + V a r ( Y ) ) ) / 2 3 resul t ing i n 18.6% ± 6 . 5 % . The variance o f the mean appears to be smaller than the i n d i v i d u a l l y l isted variances because there were numerous case studies for w h i c h variances were not g iven and assumed to be zero. The percentage o f traps lost to ghost f i sh ing is the grand mean calculated f rom case studies where the parameters were g iven, w h i c h results i n 96 .9% + 3 .1%/- 5 1 . 3 % o f lost traps (96.9% ± 51 .3%, w i t h the upper l imi t truncated at 100%). 62 The values for the l ifespan o f ghost f i sh ing traps i n this case study was an average o f thirteen other case studies, some o f w h i c h had asymmetr ica l variances. In the cases where variances were asymmetr ica l , the average o f the variances i n each case study was taken and used towards the f o l l o w i n g formula: S d ( ( X + Y ) / 1 3 ) = (sqr t (Var (X) + V a r ( Y ) ) ) / 1 3 result ing in a mean o f 0.417 + 141%/-100% years (0.417 ± 141%, w i t h the lower l imi t truncated at 0). The dis t r ibut ion o f ghost f ishing mor ta l i ty as a percentage o f the reported landed catch is shown i n F igure 3.1.24. Texas blue crab fishery 1999 or 2000 Parameter list Parameter i : 259 fishers Parameter i i : 200 traps Parameter 1: 51,800 traps Parameter 2: 4 2 . 5 % ± 18% Parameter 3: 22,105 traps Parameter 4: 96 .9% + 3.1 % / - 5 1 . 3 % Parameter 5: 21,322 traps Parameter 6: 0.417 + 141%/-100% years Parameter 7: 8,897 traps ghost fishing annually Parameter 8: 53.8 ± 3 4 . 1 % crabs- trap^year" 1 Parameter 9 : 478,635 crabs Parameter 10: 0.16 k g Parameter 11: 75,988 kg Parameter 12: 2,341,997 k g Resul t : 3.2% Narrative In the .Texas blue crab (Callinectes sapidus) fishery, 259 l icensed fishers each fished an average o f 200 traps (Hammerschmid t et al. 1998 ci ted i n G u i l l o r y et al. 2001) for a total o f 51,800 traps. The annual trap loss rate is 3 5 % - 5 0 % (Texas Parks and 63 W i l d l i f e Department unpubl ished data c i ted in G u i l l o r y et al. 2001), w h i c h corresponds to 22,015 lost traps i f the midrange were appl ied to the number o f traps used. It was unclear whether a l l lost traps were capable o f ghost f i sh ing or whether some were stolen or broken, so a modi f ie r o f 96 .9% was appl ied. T h i s modifer was the mean calculated from other case studies where this parameter was specif ied: 4 6 . 5 % i n the 1987 B C Dungeness crab fishery (Breen 1987); 100% i n the 1990 C h i n i a k B a y red k i n g crab fishery (assumed f rom R . M e y e r , pers c o m m to S m o l o w i t z 1978a); 100% i n the 1976 A t l a n t i c offshore lobster f ishery (assumed from S m o l o w i t z 1978b); 100% i n the 1969-1989 G u l f o f St. Lawrence snow crab fishery (assumed f rom V i e n n e a u and M o r i y a s u 1994); 100% i n the 1987 G u l f o f St. Lawrence snow crab fishery (Ma l l e t et al. 1998) ; 100% i n the 1987 N e w B r u n s w i c k snow crab fishery ( M a l l e t et al. 1998); 100% i n the L o u i s i a n a b lue crab fishery c. 1993 ( G u i l l o r y 1999); 100% i n the 1988 L o u i s i a n a blue crab fishery ( G u i l l o r y 1999); 100% i n the 1993 L o u i s i a n a b lue crab fishery ( G u i l l o r y 1999) ; and 100% i n the 1974 N e w f o u n d l a n d snow crab fishery ( M i l l e r 1977). Th i s results i n 21,322 lost traps capable o f ghost f ishing. There was no ind ica t ion o f h o w l o n g the lost traps w o u l d r emain funct ional , so an average length o f t ime, 0.417 years, was taken from a l l case studies i n w h i c h ghost f i sh ing pe r iod was specified. The values f rom those case studies were: 2.2 years i n the 1984 B C Dungeness crab fishery (Breen 1987); 0.418 years i n the 1969-70 A l a s k a red k i n g crab fishery ( K i m k e r 1990); 0.375 years ± 1 1 % i n the 1986 C a l i f o r n i a Dungeness crab fishery (Kennedy 1986); 0.245 ± 4 4 % / - 1 9 % years i n the 1992 C o o k Inlet red k i n g crab fishery ( K i m k e r 1990); and 0.243 ± 3 3 % years. The last value is a composi te i n itself, obtained f rom 0.241 +20.8%/-27.1% years ( K i m k e r 1990) and 0.245 +44%/-19% years ( K i m k e r 1990), and used i n 9 case studies: 1993 B e r i n g Sea crab fishery; 1993 B e r i n g Sea red k i n g crab fishery; 1990 B r i s t o l B a y red k i n g crab fishery; 1991 B r i s t o l B a y red k i n g crab fishery; 1994 B r i s t o l B a y red k i n g crab fishery; 1996 B r i s t o l B a y red k i n g crab fishery; C h i n i a k B a y red k i n g crab fishery i n the 1990s; and 2 1991 eastern B e r i n g Sea snow crab fisheries. A s s u m i n g that the rate o f entry o f lost traps was equal to the rate o f degradation, and that the numbers remained constant, 8,897 traps are estimated to be ghost f i sh ing inshore i n the area annual ly . In G u i l l o r y ' s (1993) study o f ghost f i sh ing by blue crab traps i n Lou i s i ana , an average o f 25.8 indiv iduals d ied per trap. Arcemen t and G u i l l o r y (1993) performed a compar i son o f mortal i ty o f crabs i n traps that had escape r ings and those that had no escape r ings. In the former, an average o f 5.3 indiv iduals were k i l l e d per trap i n a 2-mon th per iod. In the latter, an average o f 17.3 ind iv idua ls were k i l l e d per trap i n a 2-month per iod . A s s u m i n g that mor ta l i ty rate is un i fo rm throughout the study per iod and throughout the year, annual morta l i ty is and 31.8 and 103.8 crabs per trap respectively. T h e average o f these three values is 53.8 crabs- trap"'year"'. 64 A p p l i e d to the previous parameter, the annual morta l i ty f rom ghost traps is 478,635 crabs. The average weight o f blue crabs i n October 1999 to N o v e m b e r 2000 landings was 0.35 lbs, or 0.16 k g ( M u r p h y et al. 2001). The weight o f ghost f ishing catch o f b lue crabs is thus 75,988 k g , 3 .2% o f the reported 2,341,997 k g landed i n Texas i n 2001 (Nat iona l M a r i n e Fisher ies Serv ice 2004). Monte Carlo simulation Gear loss i n this fishery was estimated to be between 3 5 % and 5 0 % ( T P W D i n G u i l l o r y et al. 2001), w h i c h equates to 4 2 . 5 % ± 18%. The percentage o f traps lost to ghost f ishing is the grand mean calculated from case studies where the parameters were g iven , w h i c h results i n 9 6 . 9 % + 3 .1%/- 51 .3% o f lost traps (96.9% ± 5 i . 3 % , w i t h the upper l imi t truncated at 100%). The values for the l i fespan o f ghost f i sh ing traps i n this case study was an average o f thirteen other case studies, some o f w h i c h had asymmetr ica l variances. In the cases where variances were asymmetr ica l , the average o f the variances i n each case study was taken and used towards the f o l l o w i n g formula: S d ( ( X + Y ) / 1 3 ) = ( sq r t (Var (X) + V a r ( Y ) ) ) / 1 3 resul t ing i n a mean o f 0.417 + 141%/-100% years (0.417 ± 1 4 1 % , w i t h the lower l imi t truncated at 0). The mortal i ty rate is taken as the mean from three values, 53.8 crabs-trap" 1 year"1 ± 3 4 . 1 % , as expla ined i n the narrative above. The dis tr ibut ion o f ghost f i sh ing morta l i ty as a percentage o f the reported landed catch is shown i n F igure 3.1.25. Case studies with 4 missing parameters: Mississippi blue crab fishery 1999 or 2000 Parameter list Parameter i : 256 fishers Parameter i i : 346 traps per fisher Parameter 1: 88,576 traps Parameter 2: 2 5 % ± 2 0 % 65 = Parameter 3: 22,144 traps X Parameter 4: 9 6 . 9 % + 3 .1%/- 51 .3% = Parameter 5: 21,447 traps X Parameter 6: 0.417 + 141%/-100% years = Parameter 7: 8,949 traps X Parameter 8: 53.8 ± 3 4 . 1 % crabs- trap"'year = Parameter 9: 481,440 crabs X Parameter 10: 0.16 k g = Parameter 11: 75,433 kg -4- Parameter 12: 337,981 k g = Resul t : 20.2% Narrative In 1999 or 2000, 256 b lue crab (Callinectes sapidus) fishers were l icensed i n M i s s i s s i p p i ( G u i l l o r y et al. 2001) . N o estimate for the average number o f traps per fisher was avai lable, so the value for this parameter f rom the F l o r i d a fishery was appl ied instead (346 traps/fisher; A . Jackson 2004 pers. c o m m . to A . Poon) . T h i s results i n 88,576 traps used annually, assuming the numbers remain consistent. A n n u a l trap loss rate is 2 0 % -3 0 % i n M i s s i s s i p p i (T. F l o y d pers. c o m m . to G u i l l o r y et al. 2001) ; taking the midrange results i n an estimate o f 22,144 traps lost annually. It is unclear what propor t ion o f lost traps contribute to ghost f i sh ing as opposed to be ing stolen or rendered inoperable, so a modif ier o f 9 6 . 9 % was appl ied. Th i s modifer was the mean calculated f rom other case studies where this parameter was specif ied: 4 6 . 5 % i n the 1987 B C Dungeness crab fishery (Breen 1987); 100% i n the 1990 C h i n i a k B a y red k i n g crab fishery (assumed from R . M e y e r , pers c o m m to S m o l o w i t z 1978a); 100% i n the 1976 At l an t i c offshore lobster fishery (assumed f rom S m o l o w i t z 1978b); 100% i n the 1969-1989 G u l f o f St. Lawrence snow crab fishery (assumed from V i e n n e a u and M o r i y a s u 1994); 100% i n the 1987 G u l f o f St. L a w r e n c e snow crab fishery ( M a l l e t et al. 1998); 100% i n the 1987 N e w B r u n s w i c k snow crab fishery (Ma l l e t et al. 1998); 100% in the L o u i s i a n a blue crab fishery c. 1993 ( G u i l l o r y 1999); 100% i n the 1988 L o u i s i a n a blue crab fishery ( G u i l l o r y 1999); 100% i n the 1993 L o u i s i a n a blue crab fishery ( G u i l l o r y 1999); and 100% i n the 1974 N e w f o u n d l a n d snow crab fishery ( M i l l e r 1977). The number o f traps lost to ghost f ishing, then, is 21,447. In the absence o f an estimate for the functional l i fe o f the ghost traps, an average o f ghost f ishing per iod (0.417 years) was taken from a l l other case studies that had a value for this parameter was used. A s s u m i n g that the rate o f entry o f lost traps was equal to the rate o f degradation, 8,949 traps are estimated to be ghost f i sh ing i n the area annual ly . In G u i l l o r y ' s (1993) study o f ghost f i sh ing by blue crab traps i n Lou i s i ana , an 66 average o f 25.8 ind iv idua l s d ied per trap. Arcemen t and G u i l l o r y (1993) performed a compar i son o f mor ta l i ty o f crabs i n traps that had escape rings and those that had no escape rings. In the former, an average o f 5.3 ind iv idua l s were k i l l e d per trap i n a 2-month period. In the latter, an average o f 17.3 ind iv idua l s were k i l l e d per trap i n a 2-month per iod. A s s u m i n g that morta l i ty rate is un i fo rm throughout the study per iod and throughout the year, annual mortal i ty is and 31.8 and 103.8 crabs per trap respect ively. The average o f these three values is 53.8 crabs- trap"'year"'. A p p l i e d to the previous parameter, the annual mortal i ty f rom ghost traps is 481,440 crabs. The average weight o f blue crabs i n October 1999 to N o v e m b e r 2000 landings was 0.35 lbs, or 0.16 k g ( M u r p h y et al. 2001) . The weight o f ghost f i sh ing catch is thus 76,433 k g , or 2 0 . 2 % o f the reported 337,981 k g landed i n 2000 (Nat iona l M a r i n e Fisheries Service 2004) . Monte Carlo simulation F l o y d ( in G u i l l o r y et al. 2001) estimated the gear loss i n this fishery to be between 2 0 % and 3 0 % , w h i c h is a midpoin t o f 2 5 % ± 2 0 % . The percentage o f traps lost to ghost f i sh ing is the grand mean calculated f rom case studies where the parameters were g iven , w h i c h results i n 96 .9% + 3 .1%/- 51 .3% o f lost traps (96 .9% ± 51 .3%, w i t h the upper l imi t truncated at 100%). The values for the l i fespan o f ghost f i sh ing traps i n this case study was an average o f thirteen other case studies, some o f w h i c h had asymmetr ica l variances. In the cases where variances were asymmetr ica l , the average o f the variances i n each case study was taken and used towards the f o l l o w i n g formula : S d ( ( X + Y ) / 1 3 ) = ( sqr t (Var (X) + V a r ( Y ) ) ) / 1 3 result ing i n a mean o f 0.417 + 141%/-100% years (0.417 ± 141%, w i t h the l ower l i m i t truncated at 0). The morta l i ty rate is taken as the mean f rom three values, 53.8 c r abs - t r ap" ' y e a r ' ± 34 .1%, as expla ined i n the narrative above. . The dis t r ibut ion o f ghost f ishing mor ta l i ty as a percentage o f the reported landed catch is shown i n F igu re 3.1.26. . - ' '• 67 South Carolina blue crab fishery c. 1979 Parameter list Parameter i : Parameter i i : Parameter 1: Parameter 2: Parameter 3: Parameter 4: Parameter 5: Parameter 6: Parameter 7: Parameter 8: Parameter 9: Parameter 10 Parameter 11: Parameter 12 Resul t : 743 fishers 75 ± 3 3 % traps per fisher 55,725 traps 18.6% ± 6 . 5 % 10,378 traps 9 6 . 9 % + 3 .1%/- 5 1 . 3 % 10,052 traps 0.417 + 141%/-100% years 4,194 traps 53.8 ± 3 4 . 1 % crabs- trap^year" 1 225,639 crabs 0.16 k g 35,822 kg 2,980,000 k g 1.2% Narrative In the South C a r o l i n a blue crab {Callinectes sapidus) f ishery i n 1982, 743 crabbers (B i shop 1983) fished an average o f 50-100 traps (Whi taker 1979 i n G u i l l o r y 1999). I f the midrange for traps per fisher were used, 55,725 traps w o u l d be used i n total i n that fishery. N o estimate for the rate o f loss for traps was avai lable , and since no other case studies i n this paper are s imi la r enough to assume a s imi la r rate o f loss, the average annual loss rate (18.6%) f rom a l l case studies w i t h this parameter was used instead, resul t ing i n 10,378 lost traps per year. The values f rom the case studies were: 3 5 % ± 4 3 % i n the 1999/2000 A l a b a m a blue crab fishery ( L . Har tman pers. c o m m to G u i l l o r y et al. 2001) ; 2 0 % used i n the 1993 B e r i n g Sea crab fishery ( W . N i p p e s to K r u s e and K i m k e r 1993); 5% ± 3 3 % i n the 1993 B e r i n g Sea red k i n g crab fishery ( W . N i p p e s pers. c o m m . to Stevens-ef al. 1993); 10% i n the 1996 B r i s t o l B a y red k i n g crab fishery (Stevens 1996); 1% trap loss rate used i n the 1999 B e r i n g Sea snow crab fishery (R. M o r r i s o n pers. c o m m . to Stevens et al. 2000); 7 .5% ± 3 3 % i n the 1987 At l an t i c inshore lobster fishery (Krouse 1989 ci ted i n B r e e n 1990); 2 5 % ± 2 0 % i n the 1976 At l an t i c offshore lobster fishery ( S m o l o w i t z 1979b); 3 3 % i n the 1973 At l an t i c seacoast inshore lobster fishery (Prudden 1962); 10 .9% i n the 1984 B C Dungeness crab fishery (Breen 68 1987); 17.6% i n the 1975/1976 C o l u m b i a R i v e r estuary Dungeness crab fishery ( M u i r et al. 1984); 4 0 % ± 2 5 % i n the 2001 F l o r i d a b lue crab fishery ( A . M c M i l l a n - J a c k s o n pers. c o m m . to G u i l l o r y et al. 2001); 2 5 % i n the 2001 G u l f o f M e x i c o blue crab fishery ( G u i l l o r y et al. 2001); 10% i n the 1999 L o u i s i a n a blue crab fishery ( G u i l l o r y 1999); 2 2 . 5 % ± 1 1 % i n the 1975 M a i n e lobster f ishery (Sheldon and D o w 1975); 7 .5% ± 3 3 % used i n the 1992 M a i n e lobster fishery calculat ions (Carr and Har r i s 1987); 2 5 % ± 2 0 % (T. F l o y d pers. c o m m . to G u i l l o r y et al. 2001) ; 7 .5% ± 3 3 % i n the 1987 N e w E n g l a n d lobster fishery (Breen 1990); 8.3% i n the 1974 N e w f o u n d l a n d snow crab fishery (calculated f rom 1,000 traps lost f rom 12,000 vessels; M i l l e r 1977); 1 0 % i n the 1971/1972 Pac i f i c Coast Dungeness crab fishery (Paci f ic M a r i n e Fisher ies C o m m i s s i o n 1978); 12 .5% ± 2 0 % used i n the 1980 R h o d e Is land Lobster fishery (Fogarty and B o r d e n 1980 ci ted i n B r e e n 1990) 4 2 . 5 % ± 18% i n the 1999 or 2000 Texas blue crab fishery (Texas Parks and W i l d l i f e Department unpubl ished data ci ted i n G u i l l o r y et al. 2001); 11 .9% i n the Wash ing ton Dungeness crab fishery (Bar ry pers. c o m m . to B r e e n 1990); 15.7% i n the 1972 Wash ing ton Dungeness crab fishery (Tegelberg 1974); 15 .7% i n the 1973 Wash ing ton Dungeness crab fishery (Tegelberg 1974). N o t a l l lost traps are capable o f ghost f i sh ing; some are lost to vanda l i sm or are stolen, so a modi f ie r o f 96 .9% was appl ied. T h i s modifer was the mean calculated from other case studies where this parameter was specif ied: 4 6 . 5 % i n the 1987 B C Dungeness crab fishery (Breen 1987); 100% i n the 1990 C h i n i a k B a y red k i n g crab f ishery (assumed f rom R . M e y e r , pers c o m m to S m o l o w i t z 1978a); 100% i n the 1976 A t l a n t i c offshore lobster fishery (assumed f rom S m o l o w i t z 1978b); 100% i n the 1969-1989 G u l f o f St. L a w r e n c e snow crab fishery (assumed f rom V i e n n e a u and M o r i y a s u 1994); 100% i n the 1987 G u l f o f St. Lawrence snow crab fishery (Ma l l e t et al. 1998); 100% i n the 1987 N e w B r u n s w i c k snow crab fishery ( M a l l e t et al. 1998); 100% i n the L o u i s i a n a blue crab fishery c. 1993 ( G u i l l o r y 1999); 100% i n the 1988 L o u i s i a n a blue crab f ishery ( G u i l l o r y 1999); 100% i n the 1993 L o u i s i a n a blue crab fishery ( G u i l l o r y 1999); and 100% i n the 1974 N e w f o u n d l a n d snow crab fishery ( M i l l e r 1977). T h i s results i n 10,052 lost traps capable o f ghost f ishing. There was no indica t ion o f h o w long the lost traps w o u l d r emain funct ional , so an average length o f t ime, 0.417 years, was taken from a l l case studies i n w h i c h ghost f i sh ing per iod was specified. The values f rom those case studies were: 2.2 years i n the 1984 B C Dungeness crab fishery (Breen 1987); 0.418 years i n the 1969-70 A l a s k a red k i n g crab fishery ( K i m k e r 1990); 0.375 years ± 1 1 % i n the 1986 C a l i f o r n i a Dungeness crab fishery (Kennedy 1986); 0.245 +44%/-19% years i n the 1992 C o o k Inlet red k i n g crab fishery ( K i m k e r 1990); and 0.243 ± 3 3 % years. The last value is a composi te i n itself, obtained f rom 0.241 +20.8%/-27.1% years ( K i m k e r 1990) and 0.245 +44%/-19% years ( K i m k e r 1990), and used i n 9 case studies: 1993 B e r i n g Sea crab fishery; 1993 B e r i n g Sea red k i n g crab fishery; 1990 B r i s t o l B a y red k i n g crab fishery; 1991 B r i s t o l 69 B a y red k i n g crab fishery; 1994 B r i s t o l B a y red k i n g crab fishery; 1996 B r i s t o l B a y red k i n g crab fishery; C h i n i a k B a y red k i n g crab fishery i n the 1990s; and 2 1991 eastern B e r i n g Sea snow crab fisheries. A s s u m i n g that the rate o f entry o f lost traps was equal to the rate o f degradation, and that the numbers remained constant since then, 4,194 traps are estimated to be ghost f i sh ing inshore i n the area annual ly . In the absence o f an estimate for morta l i ty rate o f b lue crabs i n lost traps i n this fishery, an average f rom other case studies was used instead. In G u i l l o r y ' s (1993) study o f ghost f i sh ing b y blue crab traps i n L o u i s i a n a , an average o f 25.8 ind iv idua l s d ied per trap. Arcemen t and G u i l l o r y (1993) performed a compar i son o f mor ta l i ty o f crabs i n traps that had escape rings and those that had no escape rings: In the former, an average o f 5.3 ind iv idua l s were k i l l e d per trap i n a 2-month per iod . In the latter, an average o f 17.3 ind iv idua l s were k i l l e d per trap i n a 2-month per iod . A s s u m i n g that mor ta l i ty rate is un i fo rm throughout the study pe r iod and throughout the year, annual morta l i ty is and 31.8 and 103.8 crabs per trap respect ively. The average o f these three values is 53.8 crabs- t r ap _ 1 y e a r "" • A p p l i e d to the previous parameter, the annual mor ta l i ty f rom ghost traps is 225,639 crabs. The average weight o f b lue crabs i n October 1999 to N o v e m b e r 2000 landings was 0.35 lbs, or 0.16 k g ( M u r p h y et al. 2001). The weight o f ghost f ishing catch is thus 35,822 k g , 1.2% o f the reported landings o f 2,980,000 k g per year, averaged over 35 years (Whi taker et al. 1998). Ter rap in is a byca tch o f the blue crab fishery, and over a two month per iod , terrapin mor ta l i ty was estimated at 285 terrapins for the 743 c o m m e r c i a l crabbers i n the fishery (B i shop 1983). A s s u m i n g a constant rate o f catch, annual terrapin morta l i ty is 1,710 terrapins; a h igh number for a c losed fishery. Monte Carlo simulation , ' . Whi take r (1979 i n G u i l l o r y 1999) estimates that each fisher i n this fishery uses between 50 andlOO traps, w h i c h equates to 75 ± 3 3 % traps per fisher. The value for the trap loss rate i n this case study was the mean o f a l l other case studies w i t h this parameter. T o f ind the variance o f the mean, the f o l l o w i n g formula was used: S d ( ( X + Y ) / 2 3 ) = ( sq r t (Var (X) + Var (Y) ) ) / 23 resul t ing i n 18.6% ± 6.5%. The variance o f the mean appears to be smaller than the i n d i v i d u a l l y l isted variances because there were numerous case studies for w h i c h variances were not g iven and assumed to be zero. 70 The percentage o f traps lost to ghost f i sh ing is the grand mean calculated f rom case studies where the parameters were g iven, w h i c h results i n 96 .9% + 3 .1%/- 51 .3% o f lost traps (96.9% ± 51 .3%, w i t h the upper l imi t truncated at 100%). The values for the l ifespan o f ghost f ishing traps i n this case study was an average o f thirteen other case studies, some o f w h i c h had asymmetr ica l variances. In the cases where variances were asymmetr ica l , the average o f the variances i n each case study was taken and used towards the f o l l o w i n g formula: S d ( ( X + Y ) / 1 3 ) = ( sq r t fVar (X) + V a r ( Y ) ) ) / 1 3 resul t ing i n a mean o f 0.417 + 141%/-100% years (0.417 ± 1 4 1 % , w i t h the lower l i m i t truncated at 0). The mortal i ty rate is taken as the mean f rom three values, 53.8 crabs-trap~'year~1 ± 3 4 . 1 % , as expla ined i n the narrative above. The dis t r ibut ion o f ghost f i sh ing morta l i ty as a percentage o f the reported landed catch is shown i n F igure 3.1.27. Texas blue crab fishery c. 1997 Parameter list Parameter i : Parameter i i : Parameter 1: Parameter 2: Parameter 3: Parameter 4: Parameter 5: Parameter 6: Parameter 7: Parameter 8: Parameter 9: Parameter 10: Parameter 11: Parameter 12: Resul t : 259 fishers N o t appl icable N o t appl icable 103 traps lost per fisher 26,667 traps 9 6 . 9 % + 3 .1%/- 51 .3% 25,837 traps 0.417 + 141%/-100% years 10,781 traps 53.8 ± 3 4 . 1 % crabs- trar/ 'year" 1 579,993 crabs 0.16 k g 92,080 kg 2,341,997 k g 3.9% 71 Narrative In the Texas blue crab (Callinectes sapidus) f ishery, fishers lose an average o f 103 traps per year each (Sh ive ly 1997 ci ted i n G u i l l o r y et al. 2001). In 1999 or 2000 there were 259 l icensed fishers i n Texas ( G u i l l o r y et al. 2001), so i n that season, 26,667 traps were lost. It was unclear whether a l l lost traps were capable o f ghost f ishing or whether some were stolen or b roken , so a modi f ie r o f 9 6 . 9 % was appl ied. T h i s modifer was the mean calculated f rom other case studies where this parameter was specif ied: 4 6 . 5 % i n the 1987 B C Dungeness crab fishery (Breen 1987); 100% i n the 1990 C h i n i a k B a y red k i n g crab fishery (assumed f rom R . M e y e r , pers c o m m to S m o l o w i t z 1978a); 100% i n the 1976 At l an t i c offshore lobster fishery (assumed f rom S m o l o w i t z 1978b); 100% i n the 1969-1989 G u l f o f St. L a w r e n c e snow crab f ishery (assumed from V i e n n e a u and M o r i y a s u 1994); 100% i n the 1987 G u l f o f St. L a w r e n c e snow crab fishery ( M a l l e t et al. 1998) ; 100% i n the 1987 N e w B r u n s w i c k snow crab fishery (Ma l l e t et al. 1998); 1 0 0 % i n the L o u i s i a n a b lue crab fishery c. 1993 ( G u i l l o r y 1999); 100% i n the 1988 L o u i s i a n a b lue crab fishery ( G u i l l o r y 1999); 100% i n the 1993 L o u i s i a n a b lue crab fishery ( G u i l l o r y 1999) ; and 100% i n the 1974 N e w f o u n d l a n d snow crab fishery ( M i l l e r 1977). T h i s results i n 25,837 lost traps capable o f ghost f i sh ing. There was no ind ica t ion o f h o w long the lost traps w o u l d remain funct ional , so an average length o f t ime, 0.417 years, was taken f rom a l l case studies i n w h i c h ghost f ishing per iod was specif ied. The values f rom those case studies were: 2.2 years i n the 1984 B C Dungeness crab fishery (Breen 1987); 0.418 years i n the 1969-70 A l a s k a red k i n g crab fishery ( K i m k e r 1990); 0.375 years ± 1 1 % i n the 1986 Ca l i fo rn i a Dungeness crab fishery ( K e n n e d y 1986); 0.245 +44%/-19% years i n the 1992 C o o k Inlet red k i n g crab fishery ( K i m k e r 1990); and 0.243 ± 3 3 % years. The last value is a composi te i n itself, obtained f rom 0.241 +20.8%/-27.1% years ( K i m k e r 1990) and 0.245 +44%/-19% years ( K i m k e r 1990), and used i n 9 case studies: 1993 B e r i n g Sea crab fishery; 1993 B e r i n g Sea red k i n g crab fishery; 1990 B r i s t o l B a y red k i n g crab fishery; 1991 B r i s t o l B a y red k i n g crab fishery; 1994 B r i s t o l B a y red k i n g crab fishery; 1996 B r i s t o l B a y red k i n g crab fishery; C h i n i a k B a y red k i n g crab fishery i n the 1990s; and 2 1991 eastern B e r i n g Sea snow crab fisheries. A s s u m i n g that the rate o f entry o f lost traps was equal to the rate o f degradation, and that the numbers remained constant, 10,781 traps are estimated to be ghost f ishing inshore i n the area annual ly . In G u i l l o r y ' s (1993) study o f ghost f i sh ing by blue crab traps i n L o u i s i a n a , an average o f 25.8 ind iv idua l s d ied per trap. Arcemen t and G u i l l o r y (1993) performed a compar i son o f mor ta l i ty o f crabs i n traps that had escape rings and those that had no escape rings. In the former, an average o f 5.3 ind iv idua l s were k i l l e d per trap i n a 2-72 month per iod. In the latter, an average o f 17.3 ind iv idua l s were k i l l e d per trap i n a 2-month per iod. A s s u m i n g that mortal i ty rate is un i fo rm throughout the study pe r iod and throughout the year, annual mortal i ty is and 31.8 and 103.8 crabs per trap respect ively. The average o f these three values is 53.8 crabs- trap"'year"'. A p p l i e d to the previous parameter, the annual morta l i ty from ghost traps is 579,993 crabs. The average weight o f blue crabs i n October 1999 to N o v e m b e r 2000 landings was 0.35 lbs, or 0.16 k g ( M u r p h y et al. 2001) . The weight o f ghost f i sh ing catch o f blue crabs is thus 92,080 k g , 3 .9% o f the reported 2,341,997 k g landed i n Texas i n 2001 (Nat iona l M a r i n e Fisheries Service 2004) . Monte Carlo simulation The percentage o f traps lost to ghost f i sh ing is the grand mean calcula ted f rom case studies where the parameters were g iven , w h i c h results i n 96 .9% + 3 .1%/- 51 .3% o f lost traps (96 .9% ± 51 .3%, w i t h the upper l i m i t truncated at 100%). The values for the l ifespan o f ghost f i sh ing traps i n this case study was an average o f thirteen other case studies, some o f w h i c h had asymmetr ica l variances. In the cases where variances were asymmetr ica l , the average o f the variances i n each case study was taken and used towards the f o l l o w i n g formula: S d ( ( X + Y ) / 1 3 ) = ( sqr t (Var (X) + V a r ( Y ) ) ) / 1 3 result ing i n a mean o f 0.417 + 141%/-100% years (0.417 ± 141%, w i t h the l ower l i m i t truncated at 0). The mor ta l i ty rate is taken as the mean f rom three values, 53.8 c r abs - t r ap" ' y e a r 1 ± 3 4 . 1 % , as expla ined i n the narrative above. The dis t r ibut ion o f ghost f i sh ing mor ta l i ty as a percentage o f the reported landed catch is shown i n F igure 3.1.28. 73 Monte Carlo figures Dis t r ibu t ions o f M o n t e Ca r lo s imulat ions; x-axis denotes ghost f i sh ing as a percentage o f reported landings (%) and y-axis denotes frequency. Un les s noted otherwise, scale o f the x-ax i s begins at 0. The vert icle l ine represents the result obtained f rom the ca lcula t ion o f midrange values. 1200 i 1000 soo (300 400 200 a 6.55 7.0 7.1 7.2 7.3 7.4 7,5 7.6 Figure 3.1.1 - British Columbia Dungeness crab fishery 1984. 800 700 -300 500 •100 300 200 100 0 0 5 10 15 20 Figure 3.1.2 - California Dungeness crab fishery 1986 500 i 450 400 3S0 H 300 250 200 150 100 50 0 0.0 Figure 3 1987 0.2 0,4 0.6 0:8 1.0 .1.3 - New Brunswick snow crab fishery 700 60Q 500 400 300 200 H 100 0 0.00 0.05 0.10 0.15 0.20 Figure 3.1.4 - Gulf of St. Lawrence snow crab fishery 1966-1989 74 0,0 Figure 3.1.5 fishery 1987 0 2 0.4 0.6 '18 1,0 - Gulf of St. Lawrence snow crab 900 800 700 500 500 -I '100 300 200 100 i 0 ............ 10 15 Figure 3.1.6 - Washington State Dungeness crab fishery 1975-1976 1.0 D,S 1,0 1 5 2,0 2,5 3,0 Figure 3.1.7 - Louisiana blue crab fishery c. 1999 700 -i 600 500 400 300 200 100 0 0,0 1,0 2.0 : **•...•.. 3,0 4.0 5.0 Figure 3.1.8 - Louisiana blue crab fishery c. 1993 500 450 400 350 300 250 200 150 100 50 0 0.0 0.2 0,4 0.5 0.8 1.0 600 500 400 300 200 100 o -0.1O 0,20 Figure 3.1.9 - Louisiana blue crab fishery 1988 ;Figure 3.1.10 - Newfoundland snow crab fishery 1974 . 75 800 700 600 500 400 300 200 100 i 0 0.0 2.0 4,0 6.0 8.0 Figure 3.1.11 - Pacific Coast Dungeness crab fishery 1971-1972 900 800 700 500 J 500 400 300 200 100 0 10 12 Figure 3.1.12 - Washington State Dungeness crab fishery 1972 500 i 450 400 350 300 250 200 150 100 50 0 i 0 10 15 20 25 Figure 3.1.13 - Washington State Dungeness crab fishery 1973 700 i 600 500 ^ 400 300 200 100 0 0 10 Figure 3.1.14 - Alabama blue crab fishery 1999-2000 1000 1 900 800 700 600 500 -! 400 300 H 200 100 0 0.00 0.05 0.10 0.15 Figure 3.1.15 - Eastern Bering Sea crab fishery 1993 800 700 BOO 500 400 300 200 100 0 0,(30 0 01 0,02 0,03 0,04 0,06 Figure 3.1.16 - Eastern Bering Sea snow crab fishery 1999 76 800 700 BOO 500 -•100 300 200 100 0 0.03 0.10 D;20 0.30 0 CQ Figure 3.1.17 - Florida blue crab fishery 2001 700 -| 600 j 500 ) 400 | •* 300 1 * 200'4 * 100 >U* ••' 0 I*~-r-0 .2 8 TO Figure 3.1.18 - Gulf of Mexico blue crab fishery 1993 . r ' ' 800 700 H 600 -500 -400 300 ) 200 -j 100 ••! 0 4-0 to IS 20 Figure 3.1.19 - Gulf of Mexico blue crab fishery 1999 or 2000 TOO 900 800 700 600 500 400 300 200 100 0 0,20 0 40 mo- 0 80 Figure 3.1.20 - Gulf of St. Lawrence snow crab fishery 1980 700 •-! 600 | SOD -j 400 -| 300 -| 200 \ too J , 0 I* 0 10 20 30 40 50 500 500 •400 -| ' 300 -I 200 I ioo ; • i s *• o'I*-0:.0 0 5 1.0 1,5 2,0 2,5 3 0 Figure 3.1.21 - Louisiana blue crab fishery 1999 or Figure 3.1.22 Louisana blue crab fishery 1993 2000 77 600 S00 •100 300 i 200 100 0 0.0 0,5 1.0 1,5 2,0 2.5 700 600 500 400 300 200 100 0 0.00 0:01 0.02 0.03 0.04 0.05 Figure 3.1.23 - Ireland spider crab fishery 2000 Figure 3.1.24 - Ireland brown crab fishery 1968-1972 600 i 500 j 400 -j 300 -| i . 200 1 • 100 -j / I* 0 -r 10 12 900 800 700 600 500 400 300 200 100 0 20 40 60 80 Figure 3.1.25 - Texas blue crab fishery 1999 or Figure 3.1.26 - Mississippi blue crab fishery 1999 2000 or 2000 800 •-700 ; 600 -| 500 | 400 300 | 200 -) 100 -j 0 + 1,0 2,0 3.0 4,0 S.O 500 450 400 350 300 250 200 ISO 100 so 0 10 12 Figure 3.1.27 - South Carolina blue crab fishery Figure 3.1.28 - Texas blue crab fishery c. 1997 c. 1979 78 3.2 K i n g c r abs C a s e s tudies w i t h 1 m i s s i n g p a r a m e t e r : Chiniak Bay red king crab fishery c. 1990s Parameter list Parameter i : N o t p rov ided Parameter i i : N o t p rov ided Parameter 1: N o t p rov ided Parameter 2: N o t p rov ided Parameter 3: 3,000 traps X Parameter 4: 100% = Parameter 5: 3,000 traps X Parameter 6: 0.243 ± 3 3 % years = Parameter 7: 729 traps X Parameter 8: 0.036 crabs- trap"'year = Parameter 9: 26 crabs X Parameter 10: 0.16 k g = Parameter 11: 77 kg Parameter 12: 3,946,320 k g Resul t : 0.002% Narrative N o estimate o f f i sh ing effort or trap loss rate was g iven for the red k i n g crab (Paralithodes camtschatica) f ishery i n C h i n i a k B a y i n A l a s k a , so the total number o f traps lost annual ly from K o d i a k Is land (R. M e y e r pers. c o m m . to S m o l o w i t z 1978a) was used instead. It was not specif ied that the estimated 3,000 lost traps were capable o f ghost f i sh ing (as opposed to be ing stolen or fa l l ing to degradation) but the context i n w h i c h the estimate was g iven suggested that a l l the traps lost were ghost f i shing: " T h i s durabi l i ty made it more urgent to answer the question o f whether or not lost pots cont inued to fish and for h o w long , especia l ly when gear loss numbers were estimated at over 3,000 pots annual ly ." S ince A l a s k a n legis la t ion requires that a l l crab fisheries (aside f rom Dungeness crab fisheries) use 30 thread cotton twine (Kruse and K i m k e r 1993), the results for 30 thread twine from K i m k e r ' s (1990) study o f various sizes biodegradable twine was consul ted to estimate how long these traps w o u l d remain capable o f ghost f ishing. E i g h t repetitions i n C o o k Inlet y i e lded an average o f 89.4 days (wi th a range o f 50-106 days), or 0.245 years, to total twine degradation, and five repetitions i n Pr ince W i l l i a m Sound 79 y ie lded an average o f 87.9 days (wi th a range o f 76-99 days), or 0.241 years, for total twine degradation. N o experiment was done i n C h i n i a k B a y , so the average o f the values f rom the two locations was used instead. A s s u m i n g that the rate o f entry o f lost traps was equal to the rate o f degradation, 729 traps are estimated to be ghost f ishing i n C h i n i a k B a y that year. Several steps were taken to estimate morta l i ty i n the traps. F o u r crabs were found i n 8 pots recovered b y dragging i n a survey o f C h i n i a k B a y (Stevens 1996). S ince the species o f crab were unspecif ied, it was necessary to determine the probabi l i ty that the crabs were k i n g crabs. T o do so, catches for were obtained from the N M F S onl ine catch statistics database for a l l species o f crabs landed i n A l a s k a i n 1994. The propor t ion o f k i n g crab catches was appl ied to the 0.5 catch rate o f the traps for crabs i n C h i n i a k B a y , resul t ing i n 0.036 crabs caught per trap, or 26 for a l l traps over the year. Land ings o f red k i n g crab i n C h i n i a k B a y were unavai lable , so the closest avai lable landings - 3,946,320 k g i n K o d i a k i n the 1982/1983 season, w h i c h had the lowest landings i n 23 years ( K o d i a k Chamber o f C o m m e r c e 2001) - were used instead. T o express the number o f crabs caught i n the lost traps as a percentage o f reported landings, the average weight for red k i n g crab (6.5 lbs, or 2.95 kg) was obtained f rom K r u s e and K i m k e r (1993); the weight o f k i n g crab morta l i ty f rom ghost traps was therefore estimated to be 77 k g , or 0 .002% o f the reported landings. Monte Carlo Simulation The value for the l i fespan o f ghost f ishing traps i n this case study was an average o f two other values that had asymmetr ica l variances (0.241 years w i t h a range o f 0.208 to 0.271 years and 0.245 years w i t h a range o f 0.189 to 0.290 years). A n average was thus taken o f the variances for each case study, and the variance o f the mean was obtained f rom the f o l l o w i n g formula: S d ( ( X + Y ) / 2 ) = l / 2 s q r t ( V a r ( X ) + V a r ( Y ) ) resul t ing in a mean o f 0.243 years ± 3 3 % . The dis tr ibut ion o f ghost f i sh ing mortal i ty as a percentage o f the reported landed catch is shown i n F igure 3.2.1. 80 Case studies with 2 missing parameters: Eastern Bering Sea red king crab fishery c. 1993 Parameter list Parameter i : N o t p rov ided Parameter i i : N o t p rov ided Parameter 1: 100,000 traps X Parameter 2: 1 5 % ± 3 3 % = Parameter 3: 15,000 traps X Parameter 4: 9 6 . 9 % + 3 . 1 % / - 5 1 . 3 % Parameter 5: 14,528 traps / Parameter 6: 0.243 ± 3 3 % years = Parameter 7: 3,528 traps X Parameter 8: 1.97 ± 9 8 % crabs- trap"'year Parameter 9: 6,933 crabs Parameter 10: 2.95 k g = Parameter 11: 20,443 kg Parameter 12: 9,200,000 kg) = Resul t : 0.22% Narrative In the early 1990s, an estimated 100,000 red k i n g crab (Paralithodes camtschatica) pots were i n use i n the B e r i n g Sea (Stevens et al. 1993), w i t h an annual trap loss rate between 10 and 2 0 % ( W . N i p p e s , pers. c o m m . to Stevens et al. 1993). The midrange, 15%, is used i n the ca lcula t ion to estimate 15,000 traps lost per year i n that scenario. It is not specif ied whether the trap loss rate inc luded traps that had been stolen or were otherwise incapable o f ghost f ishing. T o achieve a conservative estimate, it was assumed that not a l l the lost traps w o u l d ghost f ish, so a modi f i e r o f 96 .9% was appl ied. T h i s modifer was the mean calculated from other case studies where this parameter was specif ied: 4 6 . 5 % i n the 1987 B C Dungeness crab fishery (Breen 1987); 100% i n the .1990 C h i n i a k B a y red k i n g crab fishery (assumed from R . M e y e r , pers c o m m to S m o l o w i t z 1978a); 100% i n the 1976 A t l a n t i c offshore lobster fishery (assumed from S m o l o w i t z 1978b); 100% i n the 1969-1989 G u l f o f St. Lawrence snow crab fishery (assumed f rom V i e n n e a u and M o r i y a s u 1994); 100% in the 1987 G u l f o f St. Lawrence snow crab fishery ( M a l l e t et al. 1998); 100% i n the 1987 N e w B r u n s w i c k snow crab fishery (Ma l l e t et al. 1998); 100% i n the L o u i s i a n a b lue crab fishery c. 1993 ( G u i l l o r y 1999); 100% i n the 1988 L o u i s i a n a blue crab fishery ( G u i l l o r y 1999); 100% i n the 1993 L o u i s i a n a b lue crab 81 fishery ( G u i l l o r y 1999); and 100% i n the 1974 N e w f o u n d l a n d snow crab fishery ( M i l l e r 1977). The result was 14,528 ghost f ishing traps. S ince A l a s k a n legis la t ion requires that a l l crab fisheries (aside f rom Dungeness crab fisheries) use 30 thread cotton twine (Kruse and K i m k e r 1993), the results for 30 thread twine from K i m k e r ' s (1990) study o f var ious sizes biodegradable twine was consulted to estimate h o w long these traps w o u l d r emain capable o f ghost f ishing. E igh t repetitions i n C o o k Inlet y i e lded an average o f 89.4 days (wi th a range o f 50-106 days), or 0.245 years, to total twine degradation, and five repetitions i n Pr ince W i l l i a m S o u n d y ie lded an average o f 87.9 days (wi th a range o f 76-99 days), or 0.241 years, to total twine degradation. N o experiment was done i n the B e r i n g Sea, so the average o f the values f rom the two locations was used instead. A s s u m i n g that the system was at equ i l ib r ium, where in ghost traps were added at the same rate as they were be ing rendered inoperable, 3,528 traps were calculated to be ghost f i sh ing annually. There was no estimate o f morta l i ty per trap i n A l a s k a ' s k i n g crab fishery, so an average was taken f rom mortal i t ies f rom regional case studies i n A l a s k a . 3.9 crabs- trap^year" 1 was f rom estimated K i m k e r (1990) and 0.03 crabs- trap"'year"' w a s estimated f rom Stevens et al. (2000) for an average o f 1.97 crabs- trap"'year"' The number o f crabs k i l l e d b y ghost traps was thus 10,083. A n estimate o f the average weight o f red k i n g crabs (6.5 lbs, or 2.95 kg) was obtained (Kruse and K i m k e r 1993); the weight o f k i n g crab mor ta l i ty from ghost traps is therefore estimated to be 29,728 k g , or 0 .22% o f the 9,200,000 k g landed i n A l a s k a i n 1993 (Reeves and Turnock 1999). Monte Carlo Simulation Gear loss for this fishery was estimated to be between 10% and 2 0 % (Stevens et al. 1993), w h i c h equates to 1 5 % ± 3 3 % . The percentage o f traps lost to ghost f i sh ing is the grand mean calculated f rom case studies where the parameters were g iven , w h i c h results i n 96 .9% + 3 .1%/- 5 1 . 3 % o f lost traps (96 .9% ± 51 .3%, w i t h the upper l i m i t truncated at 100%). The value for the l ifespan o f ghost f i sh ing traps i n this case study was an average o f two other values that had asymmetr ica l variances (0.241 years w i t h a range o f 0.208 to 0.271 years and 0.245 years w i t h a range o f 0.189 to 0.290 years). A n average was thus taken o f the variances for each case study, and the variance o f the mean was obtained f rom the f o l l o w i n g formula: S d ( ( X + Y ) / 2 ) = l / 2 s q r t ( V a r ( X ) + V a r ( Y ) ) result ing i n a mean o f 0.243 years ± 3 3 % . 82 The mortal i ty rate for this case study is taken from other A l a s k a n red k i n g crab case studies. K i m k e r (1990) estimated a morta l i ty o f 3.9 crabs- trap^year" 1 i n traps w i t h 30 twine thread, w h i l e Stevens et ah (2000) estimated 0.03 crabs per tanner crab trap. T h i s results i n an average mortal i ty rate o f 1.97 crabs per trap ± 98%. The dis t r ibut ion o f ghost f i sh ing mor ta l i ty as a percentage o f the reported landed catch is shown i n F igure 3.2.2. Bristol Bay red king crab fishery c. 1994 Parameter list Parameter i : N o t p rov ided Parameter i i : N o t p rov ided Parameter 1: N o t p rov ided Parameter 2: N o t p rov ided Parameter 3: 20,000 traps X Parameter 4: 9 6 . 9 % + 3 . 1 % / - 5 1 . 3 % = Parameter 5: 19,371 traps X Parameter 6: 0.243 ± 3 3 % years = Parameter 7: 4,705 traps X Parameter 8: 1.97 ± 98%'crabs- trap"'year = Parameter 9: 9,245 crabs X Parameter 10: 2.95 k g = Parameter 11: 27,257 kg Parameter 12: 1,025,720 k g = Resul t : 2.6% Narrative N o data on f ishing effort was g iven for one case study for the red k i n g crab (Paralithodes camtschatica) fisheries i n B r i s t o l B a y , A l a s k a , around 1994; however , the number o f lost traps was estimated to be 20,000 per year (Paul et al. 1994). It was not specif ied whether this number referred to traps capable o f ghost f i sh ing or to traps lost due to various causes, i nc lud ing those w h i c h render them incapable o f funct ioning. T o achieve a conservative estimate, it was assumed that not a l l the lost traps w o u l d ghost fish. A modi f ie r o f 96 .9% was appl ied. T h i s modifer was the mean calculated from other case studies where this parameter was specif ied: 4 6 . 5 % in the 1987 B C Dungeness crab fishery (Breen 1987); 100% i n the 1990 C h i n i a k B a y red k i n g crab fishery (assumed f rom R . M e y e r , pers c o m m to S m o l o w i t z 1978a); 100% i n the 1976 A t l a n t i c offshore lobster 83 fishery (assumed from S m o l o w i t z 1978b); 100% i n the 1969-1989 G u l f o f St. Lawrence snow crab fishery (assumed f rom V i e n n e a u and M o r i y a s u 1994); 100% i n the 1987 G u l f o f St. Lawrence snow crab fishery ( M a l l e t et al. 1998); 100% i n the 1987 N e w B r u n s w i c k snow crab fishery (Ma l l e t et al. 1998); 100% i n the L o u i s i a n a blue crab fishery c. 1993 ( G u i l l o r y 1999); 100% i n the 1988 L o u i s i a n a blue crab fishery ( G u i l l o r y 1999); 100% i n the 1993 L o u i s i a n a b lue crab fishery ( G u i l l o r y 1999); and 100% i n the 1974 N e w f o u n d l a n d snow crab fishery ( M i l l e r 1977). The result was 19,371 ghost f i sh ing traps. S ince A l a s k a n legis la t ion requires that a l l crab fisheries (aside f rom Dungeness crab fisheries) use 30-thread cotton twine (Kruse and K i m k e r 1993), the results for 30 thread twine f rom K i m k e r ' s (1990) study o f var ious sizes biodegradable twine was consul ted to estimate h o w long these traps w o u l d remain capable o f ghost f ishing. E igh t repetitions i n C o o k Inlet y i e lded an average o f 89.4 days (wi th a range o f 50-106 days), or 0.245 years, to total twine degradation, and five repetitions i n Pr ince W i l l i a m Sound y i e lded an average o f 87.9 days (wi th a range o f 76-99 days), or 0.241 years, to total twine degradation. N o experiment was done i n B r i s t o l B a y , so the average o f the values f rom the two locations was used instead. A s s u m i n g the rate o f entry o f lost traps is equal to the rate o f degradation, 4,705 traps are estimated to be ghost f i sh ing i n B r i s t o l B a y per year. N o estimate o f mor ta l i ty was g iven for crabs caught i n the traps, so an average was taken from mortal i t ies f rom other regional case studies i n A l a s k a . 3.9 crabs- trap" 'year" 1 was f rom estimated K i m k e r (1990) and 0.03 crabs- trap"'year"' was estimated from Stevens et al. (2000) for a mean o f 1.97 crabs- trap^year" 1 , g i v i n g a result o f 13,444 crabs lost to ghost f i sh ing annual ly T o express this as a percentage o f reported landings (1,025,720; Granath 2002) i n the B r i s t o l B a y fishery i n 1993, the average weight for red k i n g crab (6.5 lbs, or 2.95 kg) was obtained from K r u s e and K i m k e r (1993); the weight o f k i n g crab mortal i ty f rom ghost traps was therefore estimated to be 39,639 k g , 2 .6% o f the reported landings. Monte Carlo Simulation The percentage o f traps lost to ghost f i sh ing is the grand mean calculated from case studies where the parameters were g iven , w h i c h results i n 9 6 . 9 % + 3 .1%/- 51 .3% o f lost traps (96.9% ± 51 .3%, w i t h the upper l i m i t truncated at 100%). The value for the l i fespan o f ghost f ishing traps i n this case study was an average o f two other values that had asymmetr ica l variances (0.241 years w i t h a range o f 0.208 to 0.271 years and 0.245 years w i t h a range o f 0.189 to 0.290 years). A n average was thus taken o f the variances for each case study, and the variance o f the mean was obtained f rom the f o l l o w i n g formula: 84 S d ( ( X + Y ) / 2 ) = l / 2 s q r t ( V a r ( X ) + V a r ( Y ) ) resul t ing i n a mean o f 0.243 years ± 3 3 % . The mortal i ty rate for this case study is taken f rom other A l a s k a n red k i n g crab case studies. K i m k e r (1990) estimated a mortal i ty o f 3.9 crabs- t r a p " ' y e a r ' m t r a P s w i t h 30 twine thread, w h i l e Stevens et al. (2000) estimated 0.03 crabs per tanner crab trap. T h i s results i n an average morta l i ty rate o f 1.97 crabs per trap ± 9 8 % . The distr ibution o f ghost f i sh ing morta l i ty as a percentage o f the reported landed catch is shown i n F igure 3.2.3. Bristol Bay red king crab fishery c. 1996 Parameter list Parameter i : N o t p rov ided Parameter i i : N o t p rov ided Parameter 1: 70,000 traps X Parameter 2: 10% = Parameter 3: 7,000 traps X Parameter 4: 9 6 . 9 % + 3 . 1 % / - 5 1 . 3 % = Parameter 5: 6,780 traps ghost fishing X Parameter 6: 0.243 ± 3 3 % years = Parameter 7: 1,647 traps X Parameter 8: 1.97 ± 9 8 % crabs- trap"'year = Parameter 9: 3,236 crabs X Parameter 10: 2.95 k g = Parameter 11: 9,540 kg -r- Parameter 12: 3,812,787 k g - Resul t : 0.25% Narrative In another case study for B r i s t o l B a y , A l a s k a , Stevens (1996) used a rough approximat ion o f W . N i p p e s ' s (pers. c o m m . to Stevens et al. 1993) estimate o f 70,000 traps used annual ly i n the area's red k i n g crab (Paralithodes camtschatica) fisheries and appl ied a trap loss rate o f 10%, resul t ing i n 7,000 traps lost per year. S ince it was unclear whether this rate appl ied o n l y to traps lost i n the water and capable o f ghost f ishing, a 85 conservative approach was taken. A modi f ie r o f 9 6 . 9 % was applied. Th i s modifer was the mean calculated f rom other case studies where this parameter was specif ied: 4 6 . 5 % i n the 1987 B C Dungeness crab fishery (Breen 1987); 100% i n the 1990 C h i n i a k B a y red k i n g crab fishery (assumed f rom R . M e y e r , pers c o m m to S m o l o w i t z 1978a); 100% i n the 1976 At l an t i c offshore lobster fishery (assumed f rom S m o l o w i t z 1978b); 100% i n the 1969-1989 G u l f o f St. L a w r e n c e snow crab fishery (assumed from V i e n n e a u and M o r i y a s u 1994); 100% i n the 1987 G u l f o f St. L a w r e n c e snow crab fishery ( M a l l e t et al. 1998) ; 100% i n the 1987 N e w B r u n s w i c k snow crab fishery ( M a l l e t et al. 1998); 100% i n the L o u i s i a n a blue crab fishery c. 1993 ( G u i l l o r y 1999); 100% i n the 1988 L o u i s i a n a blue crab fishery ( G u i l l o r y 1999); 100% i n the 1993 L o u i s i a n a b lue crab fishery ( G u i l l o r y 1999) ; and 100% i n the 1974 N e w f o u n d l a n d snow crab fishery ( M i l l e r 1977). The result was 6,780 ghost f i sh ing traps. S ince A l a s k a n leg is la t ion requires that a l l crab fisheries (aside f rom Dungeness crab fisheries) use 30 thread cotton twine (Kruse and K i m k e r 1993), the results for 30 thread twine f rom K i m k e r ' s (1990) study o f var ious sizes biodegradable twine was consulted to estimate h o w long these traps w o u l d r ema in capable o f ghost f ishing. E i g h t repetitions i n C o o k Inlet y i e lded an average o f 89.4 days (wi th a range o f 50-106 days), or 0.245 years, for total twine degradation, and f ive repetitions i n Pr ince W i l l i a m Sound y ie lded an average o f 87.9 days (wi th a range o f 76-99 days), or 0.241 years, for total twine degradation. N o experiment was done i n B r i s t o l B a y , so the average o f the values from the two locations was used instead. A s s u m i n g that the rate o f entry o f lost traps was equal to the rate o f degradation, 1,647 traps are estimated to be ghost f ishing i n B r i s t o l B a y per year. N o estimate o f mor ta l i ty was g iven for crabs caught i n the traps, so an average was taken f rom mortal i t ies f rom other regional case studies i n A l a s k a . 3.9 crabs- trap" 'year" 1 was from estimated K i m k e r (1990) and 0.03 crabs- trap^year" 1 was estimated f rom Stevens et al. (2000) for a mean o f 1.97 crabs- trap"'year"', g i v i n g a result o f 4,705 crabs lost to ghost f i sh ing annual ly . T o express this as a percentage o f reported landings (3,812,787 k g ; ( A l a s k a Department o f F i s h and G a m e 2004b) i n the B r i s t o l B a y f ishery i n 1996, the average weight for red k i n g crab (6.5 lbs, or 2.95 kg) was obtained from K r u s e and K i m k e r (1993); the weight o f k i n g crab morta l i ty from ghost traps is therefore estimated to be 13,873 k g , 0 .364% o f the reported landings. Monte Carlo Simulation The percentage o f traps lost to ghost f i sh ing is the grand mean calculated f rom case studies where the parameters were g iven , w h i c h results i n 96 .9% + 3 .1%/- 5 1 . 3 % o f lost traps (96.9% ± 5 1 . 3 % , w i t h the upper l imi t truncated at 100%). 86 The value for the l ifespan o f ghost f i sh ing traps i n this case study was an average o f two other values that had asymmetr ica l variances (0.241 years w i t h a range o f 0.208 to 0.271 years and 0.245 years w i t h a range o f 0.189 to 0.290 years). A n average was thus taken o f the variances for each case study, and the variance o f the mean was obtained from the f o l l o w i n g formula : S d ( ( X + Y ) / 2 ) = l / 2 s q r t ( V a r ( X ) + V a r ( Y ) ) . result ing i n a mean o f 0.243 years ± 3 3 % . The morta l i ty rate for this case study is taken f rom other A l a s k a n red k i n g crab case studies. K i m k e r (1990) estimated a mor ta l i ty o f 3.9 crabs- trap"'year" 1 i n traps w i t h 30 twine thread, w h i l e Stevens et al. (2000) estimated 0.03 crabs per tanner crab trap. Th i s results i n an average morta l i ty rate o f 1.97 crabs per trap ± 98%. The dis t r ibut ion o f ghost f ishing mor ta l i ty as a percentage o f the reported landed catch is shown i n F igure 3.2.4. Bristol Bay red king crab fishery 1990 Parameter list Parameter i : N o t p rov ided Parameter i i : N o t p rov ided Parameter 1: 68,000 traps X Parameter 2: 2 0 % trap loss rate = Parameter 3: 7,000 traps X Parameter 4: 9 6 . 9 % + 3 . 1 % / - 5 1 . 3 % = Parameter 5: 13,172 traps X Parameter 6: 0.243 ± 3 3 % years = Parameter 7: 3,199 traps X Parameter 8: 1.97 ± 9 8 % crabs- trap"'year = Parameter 9: 6,286 crabs X Parameter 10: 2.95 k g = Parameter 11: 18,535 kg -r- Parameter 12: 9,236,358 k g = Resul t : 0.20% 87 Narrative In 1990, 68,000 traps were registered for use i n the A l a s k a B r i s t o l B a y red k i n g crab (Paralithodes camtschatica) fishery ( G r i f f i n and W a r d 1992 ci ted i n K r u s e and K i m k e r 1993). I f 2 0 % o f those traps are assumed to be lost as W . N i p p e s (pers. comm.to K r u s e and K i m k e r 1993) suggests, 13,600 traps were lost i n the fishery. It is unclear i f the trap loss rate refers sole ly to traps that end up ghost f ishing; to produce a conservative estimate, it was assumed that it d i d not. Therefore, the average o f values for other case studies where this parameter was specif ied was appl ied. Th i s modif ier , 96 .9%, was the mean calculated from other case studies where this parameter was specif ied: 4 6 . 5 % i n the 1987 B C Dungeness crab fishery (Breen 1987); 100% i n the 1990 C h i n i a k B a y red k i n g crab fishery (assumed f rom R . M e y e r , pers c o m m to S m o l o w i t z 1978a); 100% i n the 1976 At l an t i c offshore lobster fishery (assumed f rom S m o l o w i t z 1978b); 100% i n the 1969-1989 G u l f o f St. Lawrence snow crab fishery (assumed from V i e n n e a u and M o r i y a s u 1994); 100% i n the 1987 G u l f o f St. L a w r e n c e snow crab fishery ( M a l l e t et al. 1998) ; 100% i n the 1987 N e w B r u n s w i c k snow crab fishery ( M a l l e t et al. 1998); 100% i n the L o u i s i a n a b lue crab fishery c. 1993 ( G u i l l o r y 1999); 100% i n the 1988 L o u i s i a n a blue crab fishery ( G u i l l o r y 1999); 100% i n the 1993 L o u i s i a n a blue crab fishery ( G u i l l o r y 1999) ; and 100% i n the 1974 N e w f o u n d l a n d snow crab fishery ( M i l l e r 1977). T h e result was 13,172 lost traps assumed to be ghost f i sh ing. S ince A l a s k a n legis la t ion requires that a l l crab fisheries (aside f rom Dungeness crab fisheries) use 30 thread cotton twine (Kruse and K i m k e r 1993), the results for 30 thread twine f rom K i m k e r ' s (1990) study o f var ious sizes biodegradable twine was consul ted to estimate h o w long these traps w o u l d r emain capable o f ghost fishing. E i g h t repetitions i n C o o k Inlet y ie lded an average o f 89.4 days (wi th a range o f 50-106 days), or 0.245 years, to total twine degradation, and five repetitions i n Pr ince W i l l i a m Sound y ie lded an average o f 87.9 days (wi th a range o f 76-99 days), or 0.241 years, for total twine degradation. N o experiment was done i n B r i s t o l B a y , so the average o f the values f rom the two locations was used instead. A s s u m i n g that the rate o f entry o f lost traps was equal to the rate o f degradation, 3,199 traps are estimated to be ghost fishing'in B r i s t o l B a y . N o estimate o f morta l i ty was given-for crabs caught i n the traps, so an average was taken from mortali t ies from other reg ional case studies i n A l a s k a . 3.9 crabs- trap" 'year" 1 was from estimated K i m k e r (1990) and 0.03 crabs- trap"'year"' was estimated from Stevens et al. (2000) for a mean o f 1.97 crabs- trap"'year"', g i v i n g a result o f 9,142 crabs lost to ghost f i sh ing annually. T o express this as a percentage o f reported landings (9,236,358 k g ; ( A l a s k a Department o f F i s h and G a m e 2004b) i n the B r i s t o l B a y fishery i n 1990, the average weight for red k i n g crab (6.5 lbs, or 2.95 kg) was obtained f rom 88 Kruse and Kimker (1993); the weight of king crab mortality from ghost traps was therefore estimated to be 26,953 kg, or 0.292% of the reported landings. Monte Carlo Simulation The percentage of traps lost to ghost fishing is the grand mean calculated from case studies where the parameters were given, which results in 96.9% + 3.1%/- 51.3% of lost traps (96.9% ± 5.1.3%, with the upper limit truncated at 100%). The value for the lifespan of ghost fishing traps in this case study was an average of two other values that had asymmetrical variances (0.241 years with a range of 0.208 to 0.271 years and 0.245 years with a range of 0.189 to 0.290 years). An average was thus taken of the variances for each case study, and the variance of the mean was obtained from the following formula: Sd((X+Y)/2) = l/2sqrt(Var(X) + Var(Y)) resulting in a mean of 0.243 years ± 33%. The mortality rate for this case study is taken from other Alaskan red king crab case studies. Kimker (1990) estimated a mortality of 3.9 crabs- trap y^ear"1 in traps with 30 twine thread, while Stevens et al. (2000) estimated 0.03 crabs per tanner crab trap. This results in an average mortality rate of 1.97 crabs per trap ± 98%. The distribution of ghost fishing mortality as a percentage of the reported landed catch is shown in Figure 3.2.5. Bristol Bay red king crab fishery 1991 Parameter list Parameter i: Not provided Parameter ii: Not provided Parameter 1: 90,000 traps X Parameter 2: 20% = Parameter 3: 18,000 traps X Parameter 4: 96.9%+ 3.1%/-51.3% = Parameter 5: 17,434 traps X Parameter 6: 0.243 ± 33% years = Parameter 7: 4,234 traps X Parameter 8: 1.97 ± 98% crabs- trap"'year 89 X Parameter 9: Parameter 10 Parameter 11: Parameter 12 Resul t : 8,320 crabs 2.95 k g 24,531 kg 7,791,893 k g 0.32% Narrative In 1991, 90,000 traps were registered for use i n the A l a s k a B r i s t o l B a y red k i n g crab (Paralithodes camtschatica) fishery (Gr i f f i n and W a r d i n K r u s e and K i m k e r 1993). I f 2 0 % o f those traps are assumed to be lost as W . N i p p e s (pers. comm.to .Kruse and K i m k e r 1993) suggests, 18,000 traps were lost i n the fishery. T h e trap loss rate doesn' t specify whether it refers so le ly to traps that end up ghost f i sh ing; to produce a conservat ive estimate, it was assumed that it d id not. Therefore, the average o f values for other case studies where this parameter was specif ied was appl ied. Th i s modif ier , 96 .9%, was the mean calculated from other case studies where this parameter was specif ied: 4 6 . 5 % i n the 1987 B C Dungeness crab fishery (Breen 1987); 100% i n the 1990 C h i n i a k B a y red k i n g crab fishery (assumed from R . M e y e r , pe r s . comm to S m o l o w i t z 1978a); 100%) i n the 1976 At l an t i c offshore lobster fishery (assumed f rom S m o l o w i t z 1978b); 100% i n the 1969-1989 G u l f o f St. Lawrence snow crab fishery (assumed from V i e n n e a u and M o r i y a s u 1994); 100% i n the 1987 G u l f o f St. L a w r e n c e snow crab fishery ( M a l l e t et al. 1998); 100% i n the 1987 N e w B r u n s w i c k snow crab fishery ( M a l l e t et al. 1998); 100% i n the L o u i s i a n a blue crab fishery c. 1993 ( G u i l l o r y 1999); 100% i n the 1988 L o u i s i a n a blue crab fishery ( G u i l l o r y 1999); 100% i n the 1993 L o u i s i a n a blue crab fishery ( G u i l l o r y 1999); and 100% i n the 1974 N e w f o u n d l a n d snow crab fishery ( M i l l e r 1977). The result was 17,434 lost traps assumed to be ghost f ishing. S ince A l a s k a n legis la t ion requires that a l l crab fisheries (aside f rom Dungeness crab fisheries) use 30 thread cotton twine (Kruse and K i m k e r 1993), the results for 30 thread twine f rom K i m k e r ' s (1990) study o f various sizes biodegradable twine was consul ted to estimate h o w long these traps w o u l d r emain capable o f ghost f ishing. E i g h t repetitions in C o o k Inlet y i e lded an average o f 89.4 days (wi th a range o f 50-106 days), or 0.245 years, to total twine degradation, and five repetitions i n Pr ince W i l l i a m Sound y ie lded an average o f 87.9 days (wi th a range o f 76-99 days), or 0.241 years, for total twine degradation. N o experiment was done i n B r i s t o l B a y , so the average o f the values f rom the two locations was used instead. A s s u m i n g that the rate o f entry o f lost traps was equal to the rate o f degradation, 4,234 traps are estimated to be ghost f ishing i n B r i s t o l B a y . N o estimate o f mor ta l i ty was g iven for crabs caught i n the traps, so an average was taken from mortali t ies f rom other regional case studies i n A l a s k a . 3.9 crabs- trap" 90 'year" 1 was f rom estimated K i m k e r (1990) and 0.03 crabs- t rap 'year" 1 was estimated f rom Stevens et al. (2000) for a mean o f 1.97 crabs- t r a p " ' y e a r g i v i n g a result o f 12,099 crabs lost to ghost f i sh ing annual ly . T o express this as a percentage o f reported landings (7,791,893 k g ; ( A l a s k a Department o f F i s h and G a m e 2004b) i n the B r i s t o l B a y fishery i n 1991, the average weight for red k i n g crab (6.5 lbs, or 2.95 kg) was obtained from K r u s e and K i m k e r (1993); the weight o f k i n g crab morta l i ty from ghost traps is therefore estimated to be 35,673 k g , or 0 .458% o f the reported landings. Monte Carlo Simulation The percentage o f traps lost to ghost f i sh ing is the grand mean calculated f rom case studies where the parameters were g iven , w h i c h results i n 96 .9% + 3 .1%/- 5 1 . 3 % o f lost traps (96 .9% ± 51 .3%, w i t h the upper l i m i t truncated at 100%). The value for the l ifespan o f ghost f i sh ing traps i n this case study was an average o f two other values that had asymmetr ica l variances (0.241 years w i t h a range o f 0.208 to 0.271 years and 0.245 years w i t h a range o f 0.189 to 0.290 years). A n average was thus taken o f the variances for each case study, and the var iance o f the mean was obtained f rom the f o l l o w i n g formula : S d ( ( X + Y ) / 2 ) = l / 2 s q r t ( V a r ( X ) + V a r ( Y ) ) result ing i n a mean o f 0.243 years ± 3 3 % . The morta l i ty rate for this case study is taken f rom other A l a s k a n red k i n g crab case studies. K i m k e r (1990) estimated a morta l i ty o f 3.9 crabs- trap"'year" 1 i n traps w i t h 30 twine thread, w h i l e Stevens et al. (2000) estimated 0.03 crabs per tanner crab trap. T h i s results i n an average morta l i ty rate o f 1.97 crabs per trap ± 98%. The dis t r ibut ion o f ghost f ishing mor ta l i ty as a percentage o f the reported landed catch is shown i n F igure 3.2.6. Case studies with 3 missing parameters: Alaska red king crab fishery 1969-1970 Parameter list Parameter i : 354 fishers x Parameter i i : 70 traps per = Parameter 1: 24,780 traps 91 X X X X X Parameter 2: Parameter 3: Parameter 4: Parameter 5: Parameter 6: Parameter 7: Parameter 8: Parameter 9: Parameter 10 Parameter 11: Parameter 12 Resul t : 1 0 % 2,478 traps 9 6 . 9 % + 3 .1%/- 5 1 . 3 % 2,400 traps 0.418 years 1,004 traps 1.97 ± 9 8 % crabs- trap"'year 1,973 crabs 2.95 k g 5,817 kg 26,186,237 k g 0.02% .-i Narrative In the 1969-1970 season i n the A l a s k a red k i n g crab (Paralithodes camtschatica) fishery, an estimated 354 fishers fished 70 traps each (R. K a i s e r pers. c o m m . to H i g h and W o r l u n d 1979), resul t ing i n calculated total o f 24 ,780 traps. A s no estimate for trap loss rate is avai lable for A l a s k a i n general, a loss rate o f "about" 10% is appl ied f rom a case study i n the western G u l f o f A l a s k a ( H i g h 1985). The case study on ly indicated that traps were lost "as a result o f var ious mishaps" and not whether they were capable o f ghost f ishing or were s i m p l y destroyed or stolen. Therefore, o f the 2,478 traps calculated to be lost, 9 6 . 9 % o f those were assumed to be capable o f ghost f ishing. T h i s modifer was the mean calculated f rom other case studies where this parameter was specif ied: 4 6 . 5 % i n the 1987 B C Dungeness crab fishery (Breen 1987); 100% i n the 1990 C h i n i a k B a y red k i n g crab fishery (assumed f rom R . M e y e r , pers c o m m to S m o l o w i t z 1978a); 1 0 0 % i n the 1976 At l an t i c offshore lobster fishery (assumed f rom S m o l o w i t z 1978b); 100% i n the 1969-1989 G u l f o f St. Lawrence snow crab fishery (assumed from V i e n n e a u and M o r i y a s u 1994); 100% i n the 1987 G u l f o f St. L a w r e n c e snow crab fishery ( M a l l e t et al. 1998) ; 100% i n the 1987 N e w B r u n s w i c k snow crab fishery (Ma l l e t et al. 1998); 100% i n the L o u i s i a n a b lue crab fishery c. 1993 ( G u i l l o r y 1999); 100% i n the 1988 L o u i s i a n a b lue crab fishery ( G u i l l o r y 1999); 100% i n the 1993 L o u i s i a n a blue crab fishery ( G u i l l o r y 1999) ; and 100% i n the 1974 N e w f o u n d l a n d snow crab fishery ( M i l l e r 1977). W h e n that is appl ied, the number o f traps lost to ghost f i sh ing is calculated to be 2,400. A n experiment on the durabi l i ty o f crab traps us ing various biodegradable twine sizes i n Pr ince W i l l i a m Sound was done between October 1989 and M a r c h 1990. F r o m 1978 to the early 1990s, A l a s k a had a 120 thread regulat ion ( K i m k e r 1990). S ince the largest thread size used in the experiment was 96, those results (153 days, or 0.418 years, to degradation) are used for the calcula t ion o f active ghost f ishing per iod o f lost traps i n this case study. T h i s is an underestimate o f the t ime necessary for the traps to be 92 rendered inoperable, as 14 o f the 15 traps us ing 96 size twine remained intact after the study was completed. Furthermore, the 120 thread regulat ion d id not come into effect unt i l almost a decade after this case study. The number o f traps ghost f i sh ing is thus calculated to be 1,004, assuming the system is at equ i l i b r ium and that the numbers have remained constant since then. There is no estimate o f morta l i ty per trap i n A l a s k a ' s k i n g crab fishery i n general , so the average o f mortal i t ies f rom regional case studies i n A l a s k a is used instead. 3.9 crabs-trap^year" 1 was f rom estimated K i m k e r (1990) and 0.03 crabs- t r a p ~ ' y e a r ' w a s estimated f rom Stevens et al. (2000) for a mean o f 1.97 crabs- trap"'year"' The number o f crabs k i l l e d b y ghost traps is thus 2,869. A generic estimate o f the average weight o f red k i n g crabs is 6.5 lbs, or 2.95 k g (Kruse and K i m k e r 1993); the weight o f k i n g crab morta l i ty f rom ghost traps is therefore estimated to be 8,459 k g , w h i c h is 0 .032% o f the 26,186,237 k g landed i n A l a s k a i n 1969 (Nat iona l M a r i n e Fisher ies Service 2004) . Monte Carlo simulation The percentage o f traps lost to ghost f i sh ing is the grand mean calculated f rom case studies where the parameters were g iven , w h i c h results i n 9 6 . 9 % + 3 .1%/- 51 .3% o f lost traps (96 .9% ± 51 .3%, w i t h the upper l i m i t truncated at 100%). The morta l i ty rate for this case study is taken f rom other A l a s k a n red k i n g crab case studies. K i m k e r (1990) estimated a mor ta l i ty o f 3.9 crabs- trap"'year"' i n traps w i t h 30 twine thread, w h i l e Stevens et al. (2000) estimated 0.03 crabs per tanner crab trap. T h i s results i n an average morta l i ty rate o f 1.97 crabs per trap ± 98%. The dis t r ibut ion o f ghost f i sh ing mor ta l i ty as a percentage o f the reported landed catch is shown i n F igure 3.2.7. M o n t e C a r l o figures Dis t r ibut ions o f M o n t e C a r l o s imulat ions; x-axis denotes ghost f i sh ing as a percentage o f reported landings (%) and y-axis denotes frequency. Un les s noted otherwise, scale o f the x-axis begins at 0. The vert icle l ine represents the result obtained f rom the ca lcula t ion o f midrange values. 93 450 400 350 300 250 200 150 -I . 100 -|** 50 -j 0 -i — 0.000000 Figure 3.2.1 c. 1990s 0.002000 0.004000 0 0000'D 1000 -900 -BOO - * 700 -600 * 500 - • '100 300 200 - • 100 ~ 0 4 - - r 0,00 0.2.0 0.40 o.eo 0.80 Chiniak Bay red king crab fishery Figure 3.2.2 - Bristol Bay red king crab fishery c. 1993 900 j 800 i 700 j 500 ••) 500 -I '100 -| 300 4; 200 »| 100 -j o 4-' 0 900 i . 800 700 mo 500 400 300 200 100 0 0.0 0.1 0,2 0,3 0 4 0,5 0.6 0 7 Figure 3.2.3 - Bristol Bay red king crab fishery c. Figure 3.2.4 - Bristol Bay red king crab fishery 1994 1996 -300 ~| 500 j 400 j 300 -j 200 i 100 0 -h 0.0 0.2 0,4 0.6 800 700_ 500 500 400 300 200 -100 » . 0 ~ , — • • • » » * • . • 0.0 0.2 0,4 0.6 o.a 1.0 Figure 3.2.5 - Bristol Bay red king crab fishery Figure 3.2.6 - Bristol Bay red king crab fishery 1990 1991 94 600 ••) 500 i * * 400 : * * 300 -| 200 4 • - I * •mo! .* o — 0 000 0.010 0:020 0.030 0.040 0.-050' Figure 3.2.7 - Alaska red king crab fishery 1969-1970 3.3 Lobsters Case studies with 2 missing parameters: Atlantic offshore lobster fishery 1976 Parameter list Parameter i : N o t p rov ided Parameter i i : N o t p rov ided Parameter 1: 72,000 traps X Parameter 2: 2 5 % ± 2 0 % = Parameter 3: 18,000 traps X Parameter 4: 100% — Parameter 5: 18,000 traps X Parameter 6: 0.417 + 141%/-100% years = Parameter 7: 7,510 traps X Parameter 8: 14.7 ± 3 2 % lobsters- t r a p " ' y e a r = Parameter 9: 110,534 lobsters X Parameter 10: 0.585 k g = Parameter 11: 64,678 kg Parameter 12: 1,899,000 k g = Resul t : 3.4% Narrative In the At l an t i c offshore lobster {Homarus americanus) fishery, an estimated 72,000 traps were used i n 1976 ( N M F S unpubl ished data ci ted i n S m o l o w i t z 1978b), w i t h a trap loss rate o f 20 -30%. I f the median , 2 5 % , were used to calculate the number o f 95 traps lost, the result w o u l d be 18,000. S m o l o w i t z (1978b) suggests that a l l these traps were capable o f ghost f i sh ing when lost, since they were lost i n offshore grounds. There was no ind ica t ion o f h o w long the lost traps w o u l d remain functional in situ, so an average length o f t ime, 0.417 years, was taken f rom a l l case studies i n w h i c h ghost f ishing per iod was specified. The values f rom those case studies were: 2.2 years i n the 1984 B C Dungeness crab fishery (Breen 1987); 0.418 years i n the 1969-70 A l a s k a red k i n g crab fishery ( K i m k e r 1990); 0.375 years ± 1 1 % i n the 1986 C a l i f o r n i a Dungeness crab fishery (Kennedy 1986); 0.245 +44%/-19% years i n the 1992 C o o k Inlet red k i n g crab fishery ( K i m k e r 1990); and 0.243 ± 3 3 % years. The last value is a composi te i n itself, obtained f rom 0.241 +20.8%/-27.1% years ( K i m k e r 1990) and 0.245 +44%/-19% years ( K i m k e r 1990), and used i n 9 case studies: 1993 B e r i n g Sea crab fishery; 1993 B e r i n g Sea red k i n g crab fishery; 1990 B r i s t o l B a y red k i n g crab fishery; 1991 B r i s t o l B a y red k i n g crab fishery; 1994 B r i s t o l B a y red k i n g crab fishery; 1996 B r i s t o l B a y red k i n g crab fishery; C h i n i a k B a y red k i n g crab fishery i n the 1990s; and 2 1991 eastern B e r i n g Sea snow crab fisheries. A s s u m i n g that the rate o f entry o f lost traps was equal to the rate o f degradation, 7,510 traps are estimated to be ghost f i sh ing offshore i n the A t l an t i c annual ly . N o estimate o f annual morta l i ty was g iven , so an average was taken from other case studies where this parameter was available. La i s t (1996) found 24 lobsters i n l 8 pots w e i g h i n g 156.5 lbs i n total. I f each pot averaged 8.7 lbs, then it w o u l d contain 6.93 lobsters. S m o l o w i t z (1978a) estimated 0.44 lobsters- trap"'year"', Prudden (1962) estimated 48 lobsters- trap^year" 1 , a n a " R- D o w (1961 ci ted i n i n She ldon and D o w 1975) estimated 3.5 lobsters- trap^year" 1 . The mean is therefore 14.7 lobsters-trap"'year"' and the total morta l i ty from ghost f i sh ing is 110,534 lobsters per year. T o express the number o f lobsters caught i n the lost traps as a percentage o f the reported landings, the average weight for A m e r i c a n lobster (1.29 lbs, or 0.585 kg) was obtained (Merce r 2002); the weight o f lobster morta l i ty f rom ghost traps was therefore estimated to be 64,678 k g , or 3.4% o f the reported landings for the At l an t i c states (1,899,000 k g ; S m o l o w i t z 1978b). Monte Carlo simulation S m o l o w i t z (1978b) estimated the gear loss for this fishery to be between 2 0 % and 3 0 % , w h i c h equates, to a midpoin t o f 2 5 % ± 20%. The values for the l ifespan o f ghost f i sh ing traps i n this case study was an average o f thirteen other case studies, some o f w h i c h had asymmetr ica l variances. In the cases where variances were asymmetr ica l , the average o f the variances i n each case study was taken and used towards the f o l l o w i n g formula: S d ( ( X + Y ) / 1 3 ) = ( sqr t (Var (X) + V a r ( Y ) ) ) / 1 3 96 result ing i n a mean o f 0.417 + 141%/-100% years (0.417 ± 141%, w i t h the lower l imi t truncated at 0). M o r t a l i t y rate was obtained from averaging those o f other lobster case studies. La i s t (1996) estimated a morta l i ty o f 6.93 lobsters- trap"'year~ I (from 24 lobsters i n 18 pots w e i g h i n g 156.5 lbs i n total; each pot averaged 8.7 lbs). S m o l o w i t z (1978a) estimated 0.44, P rudden (1962) estimated 48, and D o w ( in She ldon and D o w 1975) estimated 3.5. The mean morta l i ty is therefore 14.7 ± 3 2 % lobsters- trap^year" 1 . The dis t r ibut ion o f ghost f ishing mor ta l i ty as a percentage o f the reported landed catch is shown i n F igure 3.3.1. Atlantic seacoast inshore lobster fishery 1973 Parameter list Parameter i : N o t p rov ided Parameter i i : N o t p rov ided Parameter 1: 2,100,000 traps Parameter 2: 3 3 % Parameter 3: 693,000 traps Parameter 4: 100% Parameter 5: 693,000 traps Parameter 6: 0.417 + 141%/-100% years Parameter 7: 289,150 traps Parameter 8: 14.7 ± 3 2 % lobsters-trap"'year Parameter 9: 4,255,571 lobsters Parameter 10: 0.585 k g Parameter 11: 2,490,122 kg Parameter 12: 11,615,400 k g Resul t : 21.4% Narrative In 1973, at least 2,100,000 lobster (Homarus americanus) traps were used inshore on the At l an t i c seacoast ( S m o l o w i t z 1978b). P rudden (1962) estimated the annual trap loss rate i n the fishery to be 3 3 % , so a total o f 693,000 traps w o u l d have been lost that year. S ince this estimate o f trap loss referred to those lost i n storms, it is assumed that a l l traps remained i n the water and were capable o f ghost f ishing. 97 Because there was no est imation o f h o w long the traps w o u l d remain capable o f ghost f i sh ing , an average (0.417 years) was taken o f the values f rom other case studies where such a number was g iven . The values from those case studies were: 2.2 years i n the 1984 B C Dungeness crab fishery (Breen 1987); 0.418 years i n the 1969-70 A l a s k a red k i n g crab fishery ( K i m k e r 1990); 0.375 years ± 1 1 % i n the 1986 C a l i f o r n i a Dungeness crab fishery (Kennedy 1986); 0.245 +44%/-19% years i n the 1992 C o o k Inlet red k i n g crab fishery ( K i m k e r 1990); and 0.243 ± 3 3 % years. T h e last value is a composi te i n itself, obtained f rom 0.241 +20.8%/-27.1% years ( K i m k e r 1990) and 0.245 +44%/-19% years ( K i m k e r 1990), and used i n 9 case studies: 1993 B e r i n g Sea crab fishery; 1993 B e r i n g Sea red k i n g crab fishery; 1990 B r i s t o l B a y red k i n g crab fishery; 1991 B r i s t o l B a y red k i n g crab fishery; 1994 B r i s t o l B a y red k i n g crab fishery; 1996 B r i s t o l B a y red k i n g crab fishery; C h i n i a k B a y red k i n g crab f ishery i n the 1990s; and 2 1991 eastern B e r i n g Sea snow crab fisheries. A s s u m i n g that the rate o f entry o f lost traps was equal to the rate o f degradation, 289,150 traps are estimated to be ghost f i sh ing offshore i n the At l an t i c inshore annual ly . N o estimate o f annual mor ta l i ty was g iven , so an average was taken from other case studies where this parameter was avai lable . La i s t (1996) found 24 lobsters i n l 8 pots w e i g h i n g 156.5 lbs i n total. I f each pot averaged 8.7 lbs, then it w o u l d contain 6.93 lobsters. S m o l o w i t z (1978a) estimated 0.44 lobsters- trap"'year" 1, P rudden (1962) estimated 48 lobsters- t r a p ~ ' y e a r " \ a n d R . D o w (1961 ci ted i n i n She ldon and D o w 1975) estimated 3.5 lobsters- trap^year" 1 . The mean is 14.7 lobsters-trap^year" 1 and the total mor ta l i ty from ghost f ishing is therefore 4,255,571 lobsters per year. T o express the number o f lobsters caught i n the lost traps as a percentage o f the reported landings, the average weight for A m e r i c a n lobster (1.29 lbs, or 0.585 kg) was obtained (Merce r 2002); the weight o f lobster mortal i ty f rom ghost traps was therefore estimated to be 2,490,122 k g , or 21 .4% o f the reported landings for the At l an t i c states (11,615,400 k g ; S m o l o w i t z 1978b). Monte Carlo simulation The values for the l ifespan o f ghost f i sh ing traps i n this case study was an average o f thirteen other case studies, some o f w h i c h had asymmetr ica l variances. In the cases where variances were asymmetr ica l , the average o f the variances i n each case study was taken and used towards the f o l l o w i n g formula: S d ( ( X + Y ) / 1 3 ) = ( sqr t fVar (X) + V a r ( Y ) ) ) / 1 3 resul t ing i n a mean o f 0.417 + 141%/-100% years (0.417 ± 141%, w i t h the lower l imi t truncated at 0). 98 M o r t a l i t y rate was obtained f rom averaging those o f other lobster case studies. L a i s t (1996) estimated a morta l i ty o f 6.93 lobsters- trap"1 year" 1 ( from 24 lobsters in 18 pots w e i g h i n g 156.5 lbs i n total; each pot averaged 8.7 lbs). S m o l o w i t z (1978a) estimated 0.44, Prudden (1962) estimated 48, and D o w ( in She ldon and D o w 1975) estimated 3.5. The mean morta l i ty is therefore 14.7 ± 3 2 % lobsters- trap"'year" 1. The distr ibution o f ghost f i sh ing morta l i ty as a percentage o f the reported landed catch is shown i n F igure 3.3.2. Maine lobster fishery 1960 Parameter list Parameter i : Parameter i i : Parameter 1: Parameter 2: Parameter 3: Parameter 4: Parameter 5: Parameter 6: Parameter 7: Parameter 8: Parameter 9: Parameter 10: Parameter 11: Parameter 12: Resul t : N o t p rov ided N o t p rov ided 745,000 traps 1 5 % ± 5 0 % trap loss rate 111,750 traps 9 6 . 9 % + 3 .1%/- 5 1 . 3 % 108,233 traps 0.417 + 141%/-100% years 45,160 traps 48 lobsters-trap"'year 1 2,167,665 lobsters 0.585 k g per lobster 1,268,396 kg 24,000,000 k g 5.3% Narrative In 1960, 745,000 traps were i n use i n the M a i n e lobster (Homarus americanus) fishery (Prudden 1962). N o estimate o f trap loss rate was avai lable for that part icular fishery, so the average o f values (15%) from other case studies i n M a i n e was used instead. Those values f rom the case studies were 7 .5% ± 3 3 % (Carr and Har r i s 1997) and 2 2 . 5 % ± 1 1 % (Sheldon and D o w 1975). T h i s results i n 111,750 traps lost annually. There is no estimate for the propor t ion o f those traps that are actual ly ghost f i sh ing , so a modif ier o f 9 6 . 9 % was appl ied. Th i s modifer was the mean calculated f rom other case studies where this parameter was specified: 4 6 . 5 % i n the 1987 B C Dungeness 99 crab fishery (Breen 1987); 100% i n the 1990 C h i n i a k B a y red k i n g crab fishery (assumed f rom R . M e y e r , pers c o m m to S m o l o w i t z 1978a); 100% i n the 1976 At l an t i c offshore lobster fishery (assumed from S m o l o w i t z 1978b); 100% i n the 1969-1989 G u l f o f St. L a w r e n c e snow crab fishery (assumed f rom V i e n n e a u and M o r i y a s u 1994); 100% i n the 1987 G u l f o f St. Lawrence snow crab fishery (Mal l e t et al. 1998); 100% i n the 1987 N e w B r u n s w i c k snow crab fishery ( M a l l e t et al. 1998); 100% i n the L o u i s i a n a blue crab fishery c. 1993 ( G u i l l o r y 1999); 100% i n the 1988 L o u i s i a n a b lue crab fishery ( G u i l l o r y 1999); 100% i n the 1993 L o u i s i a n a b lue crab fishery ( G u i l l o r y 1999); and 100% i n the 1974 N e w f o u n d l a n d snow crab fishery ( M i l l e r 1977). T h e number o f traps lost to ghost f ishing, then, is 108,233. In the absence o f an estimate for the functional l i fe o f the ghost traps, an average o f ghost f ishing per iod (0.417 years) was taken f rom a l l other case studies that had a value for this parameter was used. The values f rom those case studies were: 2.2 years i n the 1984 B C Dungeness crab fishery (Breen 1987); 0.418 years i n the 1969-70 A l a s k a red k i n g crab fishery ( K i m k e r 1990); 0.375 years ± 1 1 % i n the 1986 C a l i f o r n i a Dungeness crab fishery (Kennedy 1986); 0.245 +44%/-19% years i n the 1992 C o o k Inlet red k i n g crab fishery ( K i m k e r 1990); and 0.243 ± 3 3 % years. The last value is a composi te i n itself, obtained from 0.241 +20.8%/-27.1% years ( K i m k e r 1990) and 0.245 +44%/-19% years ( K i m k e r 1990), and used i n 9 case studies: 1993 B e r i n g Sea crab fishery; 1993 B e r i n g Sea red k i n g crab fishery; 1990 B r i s t o l B a y red k i n g crab fishery; 1991 B r i s t o l B a y red k i n g crab fishery; 1994 B r i s t o l B a y red k i n g crab fishery; 1996 B r i s t o l B a y red k i n g crab fishery; C h i n i a k B a y red k i n g crab fishery i n the 1990s; and 2 1991 eastern B e r i n g Sea snow crab fisheries. A s s u m i n g that the rate o f entry o f lost traps was equal to the rate o f degradation, 45,160 traps are estimated to be ghost f ishing i n the area annually. The annual morta l i ty per trap is 48 lobsters (Prudden 1962). The annual ghost f i sh ing morta l i ty is therefore 2,167,665 ind iv idua ls per year. T o express the number o f lobsters caught i n the lost traps as a percentage o f the reported landings, the average weight for A m e r i c a n lobster (1.29 lbs, or 0.585 kg) was obtained (Merce r 2002); the weight o f lobster mortal i ty f rom ghost traps was therefore estimated to be 1,268,396 k g , or 5 .3% o f the estimated 24,000,000 k g landed annual ly i n the At l an t i c states i n the 1960s (Prudden 1962). Monte Carlo simulation Gear loss for this case study is an average (15% ± 50%) f rom other M a i n e lobster fisheries: 7 .5% ± 3 3 % (Carr and Har r i s 1997) and 2 2 . 5 % ± 1 1 % (Sheldon and D o w 1975). S ince those case studies have their o w n uncertainties, the f o l l o w i n g formula was used to obtain the variance o f the mean: 100 S d ( ( X + Y ) / 2 ) = l / 2 s q r t ( V a r ( X ) + V a r ( Y ) ) . The percentage o f traps lost to ghost f i sh ing is the grand mean calculated f rom case studies where the parameters were g iven , w h i c h results i n 96 .9% + 3 .1%/- 5 1 . 3 % o f lost traps (96 .9% ± 51 .3%, w i t h the upper l imi t truncated at 100%). The values for the l i fespan o f ghost f i sh ing traps i n this case study was an average o f thirteen other case studies, some o f w h i c h had asymmetr ica l variances. In the cases where variances were asymmetr ica l , the average o f the variances i n each case study was taken and used towards the f o l l o w i n g formula: S d ( ( X + Y ) / 1 3 ) = ( sq r t (Var (X) + V a r ( Y ) ) ) / 1 3 result ing i n a mean o f 0.417 + 141%/-100% years (0.417 ± 141%, w i t h the l ower l i m i t truncated at 0). The dis t r ibut ion o f ghost f ishing mor ta l i ty as a percentage o f the reported landed catch is shown i n F igure 3.3.3. New England lobster fishery 1987 Parameter list Parameter i : Parameter i i : Parameter 1: Parameter 2: Parameter 3: Parameter 4: Parameter 5: Parameter 6: Parameter 7: Parameter 8: Parameter 9: Parameter 10: Parameter 11: Parameter 12: Resul t : 1.2% N o t p rov ided N o t p rov ided 1,870,000 traps 7 .5% ± 3 3 % 140,250 traps 96.9% + 3 .1%/- 51 .3% 135,836 traps 0.417 + 141%/-100% years 56,677 traps 6.93 lobsters-trap" 1 year"' 392,960 lobsters 0.585 k g 229,938 kg 18,999,188 101 Narrative In 1987, an estimated 1,870,000 lobster (Homarus americanus) traps were used i n N e w Eng land , 5%-10% o f w h i c h were lost (Krouse 1989 cited i n B r e e n 1990). I f the midrange were used to calculate the number o f traps lost, The result w a s l 4 0 , 2 5 0 traps. H o w e v e r , it is unclear what propor t ion o f those traps are capable o f ghost f i sh ing as opposed to be ing stolen or destroyed; therefore, a modi f ie r o f 96 .9% was appl ied. Th i s modifer was the mean calculated from other case studies where this parameter was specif ied: 4 6 . 5 % i n the 1987 B C Dungeness crab fishery (Breen 1987); 100% i n the 1990 C h i n i a k B a y red k i n g crab fishery (assumed from R . M e y e r , pers c o m m to S m o l o w i t z 1978a); 100% i n the 1976 At l an t i c offshore lobster fishery (assumed f rom S m o l o w i t z 1978b); 100% i n the 1969-1989 G u l f o f St. L a w r e n c e snow crab fishery (assumed from V i e n n e a u and M o r i y a s u 1994); 100% i n the 1987 G u l f o f St. Lawrence snow crab fishery (Ma l l e t et al. 1998); 100% i n the 1987 N e w B r u n s w i c k snow crab fishery ( M a l l e t et al. 1998); 100% i n the L o u i s i a n a blue crab fishery c. 1993 ( G u i l l o r y 1999); 100% i n the 1988 L o u i s i a n a b lue crab fishery ( G u i l l o r y 1999); 100% i n the 1993 L o u i s i a n a blue crab fishery ( G u i l l o r y 1999); and 100% i n the 1974 N e w f o u n d l a n d snow crab f ishery ( M i l l e r 1977). The number o f traps lost to ghost f i sh ing , then, is 135,836. In the absence o f an estimate for the funct ional l i fe o f the ghost traps, an average o f ghost f i sh ing per iod (0.417 years) was taken from a l l other case studies that had a value for this parameter was used. The values from those case studies were: 2.2 years i n the 1984 B C Dungeness crab fishery (Breen 1987); 0.418 years i n the 1969-70 A l a s k a red k i n g crab fishery ( K i m k e r 1990); 0.375 years ± 1 1 % i n the 1986 C a l i f o r n i a Dungeness crab fishery (Kennedy 1986); 0.245 ± 4 4 % / - 1 9 % years i n the 1992 C o o k Inlet red k i n g crab fishery ( K i m k e r 1990); and 0.243 ± 3 3 % years. The last va lue is a composi te i n itself, obtained from 0.241 ± 2 0 . 8 % / - 2 7 . 1 % years ( K i m k e r 1990) and 0.245 ± 4 4 % / - 1 9 % years ( K i m k e r 1990), and used i n 9 case studies: 1993 B e r i n g Sea crab fishery; 1993 B e r i n g Sea red k i n g crab fishery; 1990 B r i s t o l B a y red k i n g crab fishery; 1991 B r i s t o l B a y red k i n g crab fishery; 1994 B r i s t o l B a y red k i n g crab fishery; 1996 B r i s t o l B a y red k i n g crab fishery; C h i n i a k B a y red k i n g crab fishery i n the 1990s; and 2 1991 eastern B e r i n g Sea snow crab fisheries. A s s u m i n g that the rate o f entry o f lost traps was equal to the rate o f degradation, 56,677 traps are estimated to be ghost f i sh ing i n the area annual ly . In the N e w E n g l a n d fishery i n 1968, 24 lobsters were found i n 18 recovered pots, g i v i n g an average catch rate o f 1.33 lobsters per trap. The pots had been lost for 10 weeks. A s s u m i n g that the catch rate is un i fo rm throughout the year and is the same n o w as it was then, and be ing caught results i n morta l i ty , the mortal i ty f rom lost traps is 6.93 102 ind iv idua l s per trap per year. The annual morta l i ty from a l l ghost f i sh ing traps is thus 392,960 ind iv idua ls . T o express the number o f lobsters caught i n the lost traps as a percentage o f the reported landings, the average weight for A m e r i c a n lobster (1.29 lbs, or 0.585 kg) was obtained (Merce r 2002); the weight o f lobster morta l i ty f rom ghost traps was therefore estimated to be 229,938 k g , or 1.2% o f the estimated 18,999,188 k g landed i n the inshore and offshore lobster fisheries o f N e w E n g l a n d i n 1987 (Na t iona l M a r i n e Fisher ies Service 2004). Monte Carlo simulation The percentage o f traps lost to ghost f ishing is the grand mean calculated from case studies where the parameters were g iven , w h i c h results i n 9 6 . 9 % + 3 .1%/- 51 .3% o f lost traps (96 .9% ± 51 .3%, w i t h the upper l i m i t truncated at 100%). T h e values for the l i fespan o f ghost f i sh ing traps i n this case study was an average o f thirteen other case studies, some o f w h i c h had asymmetr ica l variances. In the cases where variances were asymmetr ica l , the average o f the variances i n each case study was taken and used towards the f o l l o w i n g formula : S d ( ( X + Y ) / 1 3 ) = ( sq r t (Var (X) + V a r ( Y ) ) ) / 1 3 resul t ing i n a mean o f 0.417 + 141%/-100% years (0.417 ± 141%, w i t h the lower l imi t truncated at 0). The dis tr ibut ion o f ghost f i sh ing mor ta l i ty as a percentage o f the reported landed catch is shown i n F igure 3.3.4. \ Case studies with 3 missing parameters: Maine lobster fishery 1975 Parameter list Parameter i : N o t p rov ided Parameter i i : N o t p rov ided Parameter 1: 1,250,000 traps X Parameter 2: 2 2 . 5 % ± 1 1 % = Parameter 3: 281,250 traps X Parameter 4: 9 6 . 9 % + 3 .1%/- 5 1 . 3 % = Parameter 5: 272,399 traps X Parameter 6: 0.417 + 141%/-100% years 103 X X Parameter 7: Parameter 8: Parameter 9: Parameter 10 Parameter 11: Parameter 12 Resul t : 113,657 traps 3.5 lobsters-trap" 1year' 397,799 lobsters 0.585 k g 232,770 kg 7,670,000 k g 3.0% - i Narrative In the early 1970s, approximate ly 1.25 m i l l i o n lobster (Homarus americanus) traps were used i n M a i n e , w i t h an average annual loss rate o f 20 to 2 5 % (Sheldon and D o w 1975). I f the midrange o f the loss rate were to be used, 281,250 traps w o u l d be lost annual ly . Howeve r , the causes o f loss inc lude storms, accidents or vanda l i sm; an estimate o f the propor t ion o f lost traps that are capable o f ghost f i sh ing is not g iven . Therefore, a modi f ie r o f 9 6 . 9 % was appl ied. Th i s modi fe r was the mean calculated from other case studies where this parameter was specified: 4 6 . 5 % i n the 1987 B C Dungeness crab fishery (Breen 1987); 100% i n the 1990 C h i n i a k B a y red k i n g crab fishery (assumed f rom R . M e y e r , pers c o m m to S m o l o w i t z 1978a); 100% i n the 1976 At l an t i c offshore lobster fishery (assumed f rom S m o l o w i t z 1978b); 100% i n the 1969-1989 G u l f o f St. L a w r e n c e snow crab fishery (assumed f rom V i e n n e a u and M o r i y a s u 1994); 100% i n the 1987 G u l f o f St. Lawrence snow crab fishery (Ma l l e t et al. 1998); 100% i n the 1987 N e w B r u n s w i c k snow crab fishery ( M a l l e t et al. 1998); 100% i n the L o u i s i a n a blue crab fishery c. 1993 ( G u i l l o r y 1999); 100% i n the 1988 L o u i s i a n a blue crab fishery ( G u i l l o r y 1999); 100% i n the 1993 L o u i s i a n a b lue crab fishery ( G u i l l o r y 1999); and 100% i n the 1974 N e w f o u n d l a n d snow crab fishery ( M i l l e r 1977). The number o f traps lost to ghost f i sh ing , then, is 272,399. In the absence o f an estimate for the functional life o f the ghost traps, an average o f ghost f i sh ing per iod (0.417 years) was taken from a l l other case studies that had a value for this parameter was used. The values from those case studies were: 2.2 years i n the 1984 B C Dungeness crab fishery (Breen 1987); 0.418 years i n the 1969-70 A l a s k a red k i n g crab fishery ( K i m k e r 1990); 0.375 years ± 1 1 % i n the 1986 C a l i f o r n i a Dungeness crab fishery ( K e n n e d y 1986); 0.245 +44%/-19% years i n the 1992 C o o k Inlet red k i n g crab fishery ( K i m k e r 1990); and 0.243 ± 3 3 % years. T h e last value is a composi te i n itself, obtained f rom 0.241 +20.8%/-27.1% years ( K i m k e r 1990) and 0.245 +44%/-19% years ( K i m k e r 1990), and used i n 9 case studies: 1993 B e r i n g Sea crab fishery; 1993 B e r i n g Sea red k i n g crab fishery; 1990 B r i s t o l B a y red k i n g crab fishery; 1991 B r i s t o l B a y red k i n g crab fishery; 1994 B r i s t o l B a y red k i n g crab fishery; 1996 B r i s t o l B a y red k i n g crab fishery; C h i n i a k B a y red k i n g crab fishery i n the 1990s; and 2 104 1991 eastern B e r i n g Sea snow crab fisheries. A s s u m i n g that the rate o f entry o f lost traps was equal to the rate o f degradation, 113,657 traps are estimated to be ghost f i sh ing i n the area annually. R . D o w (1961 ci ted i n She ldon and D o w 1975) states that lost traps catch an average o f 3.1 lobsters each; assuming that this is an annual mortal i ty , the annual morta l i ty from a l l ghost f i sh ing traps is 397,799 ind iv idua l s . T o express the number o f lobsters caught i n the lost traps as a percentage o f the reported landings, the average weight for A m e r i c a n lobster (1.29 lbs, or 0.585 kg) was obtained (Merce r 2002); the weight o f lobster morta l i ty f rom ghost traps was therefore estimated to be 232,770 k g , or 3 .0% o f the estimated 7,670,000 k g landed i n M a i n e at the t ime o f She ldon and D o w ' s (1975) publ ica t ion . Monte Carlo simulation Sheldon and D o w (1975) estimate the gear loss i n this fishery to be 20 -25%, w h i c h is 2 2 . 5 % ± 11%. The percentage o f traps lost to ghost f i sh ing is the grand mean calculated from case studies where the parameters were g iven , w h i c h results i n 96 .9% + 3 .1%/- 5 1 . 3 % o f lost traps (96.9% ± 51 .3%, w i t h the upper l imi t truncated at 100%). The values for the l i fespan o f ghost f i sh ing traps i n this case study was an average o f thirteen other case studies, some o f w h i c h had asymmetr ica l variances. In the cases where variances were asymmetr ica l , the average o f the variances i n each case study was taken and used towards the f o l l o w i n g formula: S d ( ( X + Y ) / 1 3 ) = ( sq r t fVar (X) + V a r ( Y ) ) ) / 1 3 resul t ing i n a mean o f 0.417 + 141%/-100% years (0.417 ± 141%, w i t h the lower l i m i t truncated at 0). The dis tr ibut ion o f ghost f ishing morta l i ty as a percentage o f the reported landed catch is shown i n F igure 3.3.5. Maine lobster fishery 1992 Parameter list Parameter i : N o t p rov ided Parameter i i : N o t p rov ided 105 Parameter 1: 2,000,000 traps X Parameter 2: 7 .5% ± 3 3 % = Parameter 3: 150,000 traps X Parameter 4: 9 6 . 9 % + 3.1%/- 51 .3% = Parameter 5: 145,279 traps X Parameter 6: 0 . 4 1 7 + 141%/-100% years = Parameter 7: 60,617 traps X Parameter 8: 14.7 ± 3 2 % lobsters-trap"'year = Parameter 9: 892,131 lobsters X Parameter 10: 0.585 k g = Parameter 11: 522,025kg Parameter 12: 12,170,291 = Resul t : 4.3% Narrative In 1992, an estimated 2,000,000 traps were i n use i n the M a i n e lobster (Homarus americanus) f ishery, 5%-10% o f w h i c h were lost (Krouse 1994 ci ted i n Ca r r and Har r i s 1997). I f the midrange o f the loss rate were to be used, 150,000 traps w o u l d be lost annually. H o w e v e r , the causes o f loss inc lude storms, accidents or vanda l i sm; an estimate o f the propor t ion o f lost traps that are capable o f ghost f ishing is not g iven . Therefore, a modi f i e r o f 96 .9% was appl ied. T h i s modi fe r was the mean calcula ted from other case studies where this parameter was specif ied: 4 6 . 5 % i n the 1987 B C Dungeness crab fishery (Breen 1987); 100% i n the 1990 C h i n i a k B a y red k i n g crab f ishery (assumed from R . M e y e r , pers c o m m to S m o l o w i t z 1978a); 100% i n the 1976 At l an t i c offshore lobster fishery (assumed f rom S m o l o w i t z 1978b); 100% i n the 1969-1989 G u l f o f St. Lawrence snow crab fishery (assumed f rom V i e n n e a u and M o r i y a s u 1994); 100%> i n the 1987 G u l f o f St. L a w r e n c e snow crab fishery ( M a l l e t et al. 1998); 100% i n the 1987 N e w B r u n s w i c k snow crab fishery (Mal l e t et al. 1998); 100% i n the L o u i s i a n a b lue crab fishery c. 1993 ( G u i l l o r y 1999); 100% i n the 1988 L o u i s i a n a blue crab fishery ( G u i l l o r y 1999); 100% i n the 1993 L o u i s i a n a blue crab fishery ( G u i l l o r y 1999); and 100% i n the 1974 N e w f o u n d l a n d snow crab fishery ( M i l l e r 1977). The number o f traps lost to ghost f ishing, then, is 145,279. In the absence o f an estimate for the functional l ife o f the ghost traps, an average o f ghost f i sh ing pe r iod (0.417 years) was taken f rom a l l other case studies that had a value for this parameter was used. The values f rom those case studies were: 2.2 years i n the 1984 B C Dungeness crab fishery (Breen 1987); 0.418 years i n the 1969-70 A l a s k a red k i n g crab fishery ( K i m k e r 1990); 0.375 years ± 1 1 % i n the 1986 C a l i f o r n i a Dungeness crab fishery (Kennedy 1986); 0.245 ± 4 4 % / - 1 9 % years i n the 1992 C o o k Inlet red k i n g crab fishery ( K i m k e r 1990); and 0.243 ± 3 3 % years. The last value is a 106 composi te i n itself, obtained from 0.241 +20.8%/-27.1% years ( K i m k e r 1990) and 0.245 +44%/-19% years ( K i m k e r 1990), and used i n 9 case studies: 1993 B e r i n g Sea crab fishery; 1993 B e r i n g Sea red k i n g crab fishery; 1990 B r i s t o l B a y red k i n g crab fishery; 1991 B r i s t o l B a y red k i n g crab fishery; 1994 B r i s t o l B a y red k i n g crab fishery; 1996 B r i s t o l B a y red k i n g crab fishery; C h i n i a k B a y red k i n g crab fishery i n the 1990s; and 2 1991 eastern B e r i n g Sea snow crab fisheries. A s s u m i n g that the rate o f entry o f lost traps was equal to the rate o f degradation, 60,617 traps are estimated to be ghost f i sh ing i n the area annually. N o estimate o f annual morta l i ty was g iven , so an average was taken f rom other case studies where this parameter was avai lable . La i s t (1996) found 24 lobsters i n l 8 pots w e i g h i n g 156.5 lbs i n total. I f each pot averaged 8.7 lbs, then it w o u l d conta in 6.93 lobsters. S m o l o w i t z (1978a) estimated 0.44 lobsters- trap"'year"', P rudden (1962) estimated 48 lobsters- t rap 'year" 1 , and R . D o w (1961 ci ted i n i n She ldon and D o w 1975) estimated 3.5 lobsters- trap^year" 1 . The mean is thus 14.7 lobsters-trap~ 1year" 1 and the total annual mor ta l i ty from a l l ghost f i sh ing traps is 892,131 ind iv idua l s . T o express the number o f lobsters caught i n the lost traps as a percentage o f the reported landings, the average weight for A m e r i c a n lobster (1.29 lbs, or 0.585 kg) was obtained ( M e r c e r 2002); the weight o f lobster mortal i ty f rom ghost traps was therefore estimated to be 522,025 k g , or 4 . 3 % o f the 12,170,291 k g landed i n 1992 i n the inshore and offshore fisheries o f M a i n e (Na t iona l M a r i n e Fisher ies Service 2004) . Monte Carlo simulation C a r r and Har r i s (1997) estimate the gear loss i n this fishery at 5%-10%, w h i c h is 7 .5% ± 3 3 % . The percentage o f traps lost to ghost f i sh ing is the grand mean calcula ted f rom case studies where the parameters were g iven , w h i c h results i n 9 6 . 9 % + 3 .1%/ - 51 .3% o f lost traps (96 .9% ± 51 .3%, w i t h the upper l i m i t truncated at 100%). The values for the l ifespan o f ghost f i sh ing traps i n this case study was an average o f thirteen other case studies, some o f w h i c h had asymmetr ica l variances. In the cases where variances were asymmetr ica l , the average o f the variances i n each case study was taken and used towards the f o l l o w i n g formula : 1 S d ( ( X + Y ) / 1 3 ) = ( sqr t (Var (X) + V a r ( Y ) ) ) / 1 3 resul t ing i n a mean o f 0.417 + 141%/-100% years (0.417 ± 141%, w i t h the l ower l imi t truncated at 0). 107 M o r t a l i t y rate was obtained f rom averaging those o f other lobster case studies. La i s t (1996) estimated a morta l i ty o f 6.93 lobsters- trap"'year"' ( from 24 lobsters i n 18 pots w e i g h i n g 156.5 lbs i n total; each pot averaged 8.7 lbs). S m o l o w i t z (1978a) estimated 0.44, P rudden (1962) estimated 48, and D o w ( in She ldon and D o w 1975) estimated 3.5. The mean morta l i ty is therefore 14.7 ± 3 2 % lobsters- t r a p " ' y e a r The dis t r ibut ion o f ghost f ishing mor ta l i ty as a percentage o f the reported landed catch is shown i n F igure 3.3.6. Case studies with 4 missing parameters: Maine and Massachusetts lobster fishery 1991 Parameter list Parameter i : N o t p rov ided Parameter i i : N o t p rov ided Parameter 1: 2,500,924 traps X Parameter 2: 1 5 % ± 5 0 % = Parameter 3: 375,139 traps X Parameter 4: 9 6 . 9 % + 3 . 1 % / - 5 1 . 3 % = Parameter 5: 363,333 traps X Parameter 6: 0.417 + 141%/-100% years = Parameter 7: 151,599 traps X Parameter 8: XA.l ± 3 2 % lobsters-trap"'year = Parameter 9: 2,231,152 lobsters X Parameter 10: 0.585 k g = Parameter 11: 1,305,545 kg Parameter 12: 20,920,780 k g = Resul t : 6.2% Narrative In 1991, 2,500,924 lobster (Homarus americanus) traps were used i n Massachusetts and M a i n e (Carr and Har r i s 1997). N o estimate o f trap loss rate was avai lable for that part icular fishery, so the average o f values (15%) from other case studies i n M a i n e was used instead. Those values f rom the case studies were 7 .5% ± 3 3 % (Carr and Har r i s 1997) and 2 2 . 5 % ± 1 1 % (She ldon and D o w 1975). Therefore, 375,139 traps w o u l d be lost annual ly . 108 H o w e v e r , the causes o f loss inc lude storms, accidents or vanda l i sm; an estimate o f the propor t ion o f lost traps that are capable o f ghost f ishing is not g iven . Therefore, a modi f ie r o f 9 6 . 9 % was applied. Th i s modifer was the mean calculated from other case studies where this parameter was specif ied: 4 6 . 5 % i n the 1987 B C Dungeness crab fishery (Breen 1987); 100% i n the 1990 C h i n i a k B a y red k i n g crab fishery (assumed f rom R . M e y e r , pers c o m m to S m o l o w i t z 1978a); 100% i n the 1976 At l an t i c offshore lobster fishery (assumed f rom S m o l o w i t z 1978b); 100% i n the 1969-1989 G u l f o f St. Lawrence snow crab fishery (assumed f rom V i e n n e a u and M o r i y a s u 1994); 100% i n the 1987 G u l f " o f St. L a w r e n c e snow crab fishery ( M a l l e t et al. 1998); 100% i n the 1987 N e w B r u n s w i c k snow crab fishery (Mal l e t et al. 1998); 100% i n the L o u i s i a n a blue crab fishery c. 1993 ( G u i l l o r y 1999); 100% i n the 1988 L o u i s i a n a blue crab fishery ( G u i l l o r y 1999); 100% i n the 1993 L o u i s i a n a blue crab fishery ( G u i l l o r y 1999); and 100% i n the 1974 N e w f o u n d l a n d snow crab fishery ( M i l l e r 1977). The number o f traps lost to ghost f ishing, then, is 363,333. In the absence o f an estimate for the functional l i fe o f the ghost traps, an average o f ghost f i sh ing pe r iod (0.417 years) was taken f rom a l l other case studies that had a value for this parameter was used. The values from those case studies were: 2.2 years i n the 1984 B C Dungeness crab fishery (Breen 1987); 0.418 years i n the 1969-70 A l a s k a red k i n g crab fishery ( K i m k e r 1990); 0.375 years ± 1 1 % i n the 1986 C a l i f o r n i a Dungeness crab fishery (Kennedy 1986); 0.245 +44%/-19% years i n the 1992 C o o k Inlet red k i n g crab fishery ( K i m k e r 1990); and 0.243 ± 3 3 % years. The last value is a composi te i n itself, obtained from 0.241 +20.8%/-27.1% years ( K i m k e r 1990) and 0.245 +44%/-19% years ( K i m k e r 1990), and used i n 9 case studies: 1993 B e r i n g Sea crab fishery; 1993 B e r i n g Sea red k i n g crab fishery; 1990 B r i s t o l B a y red k i n g crab fishery; 1991 B r i s t o l B a y red k i n g crab fishery; 1994 B r i s t o l B a y red k i n g crab fishery; 1996 B r i s t o l B a y red k i n g crab fishery; C h i n i a k B a y red k i n g crab fishery i n the 1990s; and 2 1991 eastern B e r i n g Sea snow crab fisheries. A s s u m i n g that the rate o f entry o f lost traps was equal to the rate o f degradation, 15.1,599 traps are estimated to be ghost f i sh ing i n the area annually. . N o estimate o f annual morta l i ty was g iven , so an average was taken f rom other case studies where this parameter was available. La i s t (1996) found 24 lobsters i n l 8 pots w e i g h i n g 156.5 lbs i n total. I f each pot averaged 8.7 lbs, then it w o u l d conta in 6.93 lobsters. S m o l o w i t z (1978a) estimated 0.44 lobsters- trap"'year"', P rudden (1962) estimated 48 lobsters- trap"'year"', and R . D o w (1961 ci ted i n i n She ldon and D o w 1975) estimated 3.5 lobsters- trap"'year"'. The mean is thus 14.7 lobsterstrap"'year" 1 and the total mor ta l i ty is 2,231,152 ind iv idua ls . T o express the number o f lobsters caught i n the lost traps as a percentage o f the reported landings, the average weight for A m e r i c a n lobster (1.29 lbs, or 0.585 kg) was obtained (Merce r 2002); the weight o f lobster mor ta l i ty f rom ghost traps was therefore estimated to be 1,305,545 k g , or 6 .2% o f the 109 estimated 20,920,780 k g landed i n the inshore and offshore lobster fisheries o f M a i n e and Massachuset ts i n 1991. Monte Carlo simulation Gear loss for this case study is an average (15% ± 50%) f rom other M a i n e lobster fisheries: 7 .5% ± 3 3 % (Carr and Har r i s 1997) and 2 2 . 5 % ± 1 1 % (She ldon and D o w 1975). S ince those case studies have their o w n uncertainties, the f o l l o w i n g formula was used to obtain the variance o f the mean: S d ( ( X + Y ) / 2 ) = l / 2 s q r t ( V a r ( X ) + V a r ( Y ) ) . The percentage o f traps lost to ghost f ishing is the grand mean calculated from case studies where the parameters were g iven , w h i c h results i n 9 6 . 9 % + 3 .1%/- 51 .3% o f lost traps (96 .9% ± 51 .3%, w i t h the upper l i m i t truncated at 100%). The values for the l i fespan o f ghost f i sh ing traps i n this case study was an average o f thirteen other case studies, some o f w h i c h had asymmetr ica l variances. In the cases where variances were asymmetr ica l , the average o f the variances i n each case study was taken and used towards the f o l l o w i n g formula : S d ( ( X + Y ) / 1 3 ) = ( sq r t fVar (X) + V a r ( Y ) ) ) / 1 3 resul t ing i n a mean o f 0.417 + 141%/-100% years (0.417 ± 1 4 1 % , w i t h the lower l imi t truncated at 0). M o r t a l i t y rate was obtained f rom averaging those o f other lobster case studies. La i s t (1996) estimated a morta l i ty o f 6.93 lobsters- trap^year" 1 ( from 24 lobsters i n 18 pots w e i g h i n g 156.5 lbs i n total; each pot averaged 8.7 lbs). S m o l o w i t z (1978a) estimated 0.44, Prudden (1962) estimated 48, and D o w ( in She ldon and D o w 1975) estimated 3.5. The mean morta l i ty is therefore 14.7 ± 3 2 % lobsters- trap^year" 1 . The dis tr ibut ion o f ghost f i sh ing morta l i ty as a percentage o f the reported landed catch is shown in F igure 3.3.7. Atlantic lobster fishery 1988 Parameter list Parameter i : N o t p rov ided Parameter i i : N o t p rov ided 110 Parameter 1: N o t p rov ided Parameter 2: N o t p rov ided Parameter 3: 500,000 traps X Parameter 4: 9 6 . 9 % + 3 .1%/- 51 .3% - Parameter 5: 484,265 traps X Parameter 6: 0.417 + 141%/-100% years = Parameter 7: 202,057 traps X Parameter 8: 14.7 ± 3 2 % lobsters-trap"'year = Parameter 9: 2,973,771 lobsters X Parameter 10: 0.585 k g = Parameter 11: 1,740,084 kg Parameter 12: 24,853,304 k g = Resul t : 7.0% Narrative A n unpubl ished study b y the Center for Env i ronmen ta l Educa t ion (1987 ci ted i n B r e e n 1990) estimated an annual trap loss o f 500,000 i n the At l an t i c lobster (Homarus americanus) fishery. It was not specif ied whether the trap loss rate referred to traps capable o f ghost f i sh ing or to traps lost due to var ious causes, i nc lud ing those that render them incapable o f funct ioning. T o achieve a conservative estimate, it was assumed that not a l l the lost traps w o u l d ghost fish. A modif ie r o f 9 6 . 9 % was appl ied. T h i s modi fe r was the mean calculated f rom other case studies where this parameter was specif ied: 4 6 . 5 % i n the 1987 B C Dungeness crab fishery (Breen 1987); 100% i n the 1990 C h i n i a k B a y red k i n g crab fishery (assumed from R . M e y e r , pers c o m m to S m o l o w i t z 1978a); 100% i n the 1976 At l an t i c offshore lobster fishery (assumed f rom S m o l o w i t z 1978b); 100% i n the 1969-1989 G u l f o f St. Lawrence snow crab fishery (assumed from V i e n n e a u and M o r i y a s u 1994); 100% i n the 1987 G u l f o f St. L a w r e n c e snow crab fishery ( M a l l e t et al. 1998); 100% i n the 1987 N e w B r u n s w i c k snow crab fishery (Ma l l e t et al. 1998); 100% i n the L o u i s i a n a blue crab fishery c. 1993 ( G u i l l o r y 1999); 100% in the 1988 L o u i s i a n a blue crab f ishery ( G u i l l o r y 1999); 100% i n the 1993 L o u i s i a n a blue crab fishery ( G u i l l o r y 1999); and 100% i n the 1974 N e w f o u n d l a n d snow crab fishery ( M i l l e r 1977). The result was 484,265 ghost fish traps. There was no ind ica t ion o f h o w long the lost traps w o u l d remain functional in situ, so an average length o f t ime, 0.417 years, was taken f rom a l l case studies i n w h i c h ghost f ishing per iod was specif ied. The values f rom those case studies were: 2.2 years i n the 1984 B C Dungeness crab fishery (Breen 1987); 0.418 years i n the 1969-70 A l a s k a red k i n g crab fishery ( K i m k e r 1990); 0.375 years ± 1 1 % i n the 1986 Ca l i fo rn i a Dungeness crab fishery ( K e n n e d y 1986); 0.245 +44%/-19% years i n the 1992 C o o k Inlet red k i n g crab fishery ( K i m k e r 1990); and 0.243 ± 3 3 % years. T h e last value is a 111 composi te i n itself, obtained from 0.241 +20.8%/-27.1% years ( K i m k e r 1990) and 0.245 +44%/-19% years ( K i m k e r 1990), and used i n 9 case studies: 1993 B e r i n g Sea crab fishery; 1993 B e r i n g Sea red k i n g crab fishery; 1990 B r i s t o l B a y red k i n g crab fishery; 1991 Br i s t o l B a y red k i n g crab fishery; 1994 B r i s t o l B a y red k i n g crab fishery; 1996 B r i s t o l B a y red k i n g crab fishery; C h i n i a k B a y red k i n g crab fishery i n the 1990s; and 2 1991 eastern B e r i n g Sea snow crab fisheries. A s s u m i n g that the rate o f entry o f lost traps was equal to the rate o f degradation, 202,057 traps are estimated to be ghost f ishing inshore i n the At l an t i c annually. N o estimate o f annual mortal i ty was g iven , so an average (14.7 lobsters-trap"'year" ') was taken from other case studies where this parameter was available. La i s t (1996) found 24 lobsters i n l 8 pots w e i g h i n g 156.5 lbs i n total. I f each pot averaged 8.7 lbs, then it w o u l d contain 6.93 lobsters. S m o l o w i t z (1978a) estimated 0.44 lobsters- trap"'year"', Prudden (1962) estimated 48 lobsters- trap"'year"', and R . D o w (1961 ci ted i n i n She ldon and D o w 1975) estimated 3.5 lobsters- t rap 'year" ' . The mean is therefore 14.7 lobsters-trap"'year"' and the total mortal i ty f rom ghost f ishing is 2,973,771 lobsters per year. T o express the number o f lobsters caught i n the lost traps as a percentage o f the reported landings, the average weight for A m e r i c a n lobster (1.29 lbs, or 0.585 kg) was obtained (Merce r 2002); the weight o f lobster morta l i ty from ghost traps was therefore estimated to be 1,740,084 k g , or 7.0% o f the reported landings for the At lan t i c inshore i n 2001 (24,853,304 k g ; N a t i o n a l M a r i n e Fisheries Service 2004). Monte Carlo simulation The percentage o f traps lost to ghost f ishing is the grand mean calculated f rom case studies where the parameters were g iven , w h i c h results i n 96 .9% + 3 .1%/- 51 .3% o f lost traps (96.9% ± 51 .3%, w i t h the upper l imi t truncated at 100%). The values for the l ifespan o f ghost f ishing traps i n this case study was an average o f thirteen other case studies, some o f w h i c h had asymmetr ica l variances. In the cases where variances were asymmetr ica l , the average o f the variances i n each case study was taken and used towards the f o l l o w i n g formula: S d ( ( X + Y ) / 1 3 ) = ( sqr t (Var (X) + Var (Y) ) ) / 13 resulting i n a mean o f 0.417 + 141%/-100% years (0.417 ± 141%, w i t h the lower l imi t truncated at 0). M o r t a l i t y rate was obtained from averaging those o f other lobster case studies. La i s t (1996) estimated a morta l i ty o f 6.93 lobsters- trap"'year"' (from 24 lobsters i n 18 pots we igh ing 156.5 lbs i n total; each pot averaged 8.7 lbs). S m o l o w i t z (1978a) 112 estimated 0.44, Prudden (1962) estimated 48, and Dow (in Sheldon and Dow 1975) estimated 3.5. The mean mortality is therefore 14.7 ± 32% lobsters- trap"'year"'. The distribution of ghost fishing mortality as a percentage of the reported landed catch is shown in Figure 3.3.8. Monte Carlo figures Distributions of Monte Carlo simulations; x-axis denotes ghost fishing as a percentage of reported landings (%) and y-axis denotes frequency. Unless noted otherwise, scale of the x-axis begins at 0. The verticle line represents the result obtained from the calculation of midrange values. 500 -. 400 3DD 200 100 0 * _ • * * • * * * *• 4 * • *•** * 2 to Figure 3.3.1 - Atlantic offshore lobster fishery 1976 800 7CO H 6G0 SCO 4GO 3G0 200 1CO 0 20 40 Figure 3.3.2 - Atlantic seacoast inshore lobster fishery 1973 900 800 700 500 500 400 300 200 100 0 10 15 Figure 3.3.3 - Maine lobster fishery 1960 800 700 -300 500 400 -j 300 200 100 0 0.0 0,5 1.0 1.5 2.0 25 3.0 Figure 3.3.4 - New England lobster fishery 1987 113 - * * • • 1 + • * * * * •* • + * * * * • + I • — • 1 1 * * * 1— . . i 0 2 4 6 Figure 3.3.5 - Maine lobster fishery 1975 1400 1200 1000 800 600 400 200 0 4 0 10 20 30 Figure 3.3.6 - Maine lobster fishery 1992 40 1000 900 800 700 600 500 400 300 200 100 0 **. • • * * • .«-• <* *• ******, 20 40 Figure 3.3.7 - Maine and Massachusetts lobster fishery 1991 700 600 500 400 i 300 200 100 0 60 0 5 10 15 20 Figure 3.3.8 - Atlantic lobster fishery 1988 25 3.4 Spiny lobsters Case studies with 3 missing parameters Hawaiian spiny lobster fishery 1989 Parameter list Parameter i : Not provided Parameter i i : Not provided Parameter 1: 1,000,000 traps Parameter 2: Not provided Parameter 3: 1,084 traps lost Parameter 4: 96.9% + 3.1 %/- 51.3% Parameter 5: 1,050 traps Parameter 6: 0.417 4- 141 %/-100% years Parameter 7: 438 traps Parameter 8: 0.304 lobsters-trap"'year"' 114 = Parameter 9: 133 lobsters x Parameter 10: 0.47 +185%/-74% k g per lobster = Parameter 11: 62.1 kg + Parameter 12: 261,274 k g Resul t : 0.024% Narrative In 1989, more than 1,000,000 spiny lobster (Panulirus marginatus) traps were set i n the Nor thwestern H a w a i i a n Islands (Parr ish and K a z a m a 1992), and 1,084 traps were lost (Landgra f et al. 1990). T h i s estimate o f trap loss m a y be an underestimate, since it is h i g h l y u n l i k e l y that the rate o f trap loss cou ld be so sma l l compared to the number o f traps. Because there was no ind ica t ion whether a l l the lost traps were lost to ghost f i sh ing , a modi f ie r o f 9 6 . 9 % was appl ied. T h i s modifer was the mean calculated f rom other case studies where this parameter was specif ied: 4 6 . 5 % i n the 1987 B C Dungeness crab fishery (Breen 1987); 100% i n the 1990 C h i n i a k B a y red k i n g crab fishery (assumed f rom R . M e y e r , pers c o m m to S m o l o w i t z 1978a); 100% i n the 1976 At l an t i c offshore lobster f ishery (assumed f rom S m o l o w i t z 1978b); 100% i n the 1969-1989 G u l f o f St. L a w r e n c e snow crab fishery (assumed f rom V i e n n e a u and M o r i y a s u 1994); 100% i n the 1987 G u l f o f St. Lawrence snow crab fishery (Ma l l e t et al. 1998); 100% i n the 1987 N e w B r u n s w i c k snow crab fishery ( M a l l e t et al. 1998); 100% i n the L o u i s i a n a blue crab fishery c. 1993 ( G u i l l o r y 1999); 100% i n the 1988 L o u i s i a n a b lue crab fishery ( G u i l l o r y 1999); 100% i n the 1993 L o u i s i a n a b lue crab fishery ( G u i l l o r y 1999); and 100% i n the 1974 N e w f o u n d l a n d snow crab fishery ( M i l l e r 1977). T h i s results i n 1,050 traps capable o f ghost f ishing. N o estimate for the functional l ifespan o f the traps was avai lable , so an average length o f t ime, 0.417 years, was taken f rom a l l case studies i n w h i c h an estimate o f ghost f i sh ing per iod was g iven, and appl ied to the ghost f i sh ing traps. The values f rom those case studies were: 2.2 years i n the 1984 B C Dungeness crab fishery (Breen 1987); 0.418 years i n the 1969-70 A l a s k a red k i n g crab fishery ( K i m k e r 1990); 0.375 years ± 1 1 % i n the 1986 Ca l i fo rn i a Dungeness crab fishery (Kennedy 1986); 0.245 +44%/-19% years i n the 1992 C o o k Inlet red k i n g crab fishery ( K i m k e r 1990); and 0.243 ± 3 3 % years. The last va lue is a composite i n itself, obtained f rom 0.241 +20.8%/-27.1% years ( K i m k e r 1990) and 0.245 +44%/-19% years ( K i m k e r 1990), and used i n 9 case studies: 1993 B e r i n g Sea crab fishery; 1993 B e r i n g Sea red k i n g crab fishery; 1990 B r i s t o l B a y red k i n g crab fishery; 1991 B r i s t o l B a y red k i n g crab fishery; 1994 B r i s t o l B a y red k i n g crab fishery; 1996 B r i s t o l B a y red k i n g crab fishery; C h i n i a k B a y red k i n g crab fishery i n the 1990s; and 2 1991 eastern B e r i n g ' S e a snow crab fisheries. A s s u m i n g that the ./rate o f 115 entry o f lost traps was equal to the rate o f degradation, 438 traps are estimated to be ghost f i sh ing i n the area annually. Par r i sh and K a z a m a (1992) state that morta l i ty f rom ghost f i sh ing traps is 0.22 that o f ful ly funct ioning traps, so obtain an estimate o f the mor ta l i ty per ghost trap, 0.22 o f the catch per unit effort i n 1989 (1.38; L a n d g r a f et al. 1990) was taken and d iv ided by the number o f traps set, resul t ing i n a morta l i ty o f 0.304 ind iv idua l s per trap. The number o f lobsters lost to ghost f i sh ing is thus 133 per year. The average weight o f sp iny lobsters is 0.47 k g , taken f rom a mean carapace length o f 80.9 m m ( D e M a r t i n i et al. 2003) and a bodyweigh t to carapace length mode l o f BW=0.00090CL19952 ( U c h i d a and T a g a m i 1984 ci ted i n D e M a r t i n i et al. 2003). The weight o f catches f rom ghost f i sh ing traps is thus 62.1 k g , or 0 .024% o f the reported landings for the Nor thwestern H a w a i i a n Islands i n 1989 (261,274 k g ; L a n d g r a f et al. 1990). Monte Carlo simulation The percentage o f traps lost to ghost f i sh ing is the grand mean calculated from case studies where the parameters were g iven , w h i c h results i n 9 6 . 9 % + 3 .1%/- 51 .3% o f lost traps (96.9% ± 51 .3%, w i t h the upper l imi t truncated at 100%). The values for the l i fespan o f ghost f ishing traps i n this case study was an average o f thirteen other case studies, some o f w h i c h had asymmetr ica l variances. In the cases where variances were asymmetr ica l , the average o f the variances i n each case study was taken and used towards the f o l l o w i n g formula: S d ( ( X + Y ) / l 3) = ( sq r t (Var (X) + V a r ( Y ) ) ) / l 3 resul t ing i n a mean o f 0.417 + 141%/-100% years (0.417 ± 1 4 1 % , w i t h the lower l imi t truncated at 0). The values for sp iny lobster weight were obtained f rom D e M a r t i n i et al. 's (2003) study, where the carapace lengths ranged f rom 51.9 m m to 114.7 m m w i t h a mean o f 80.9 m m . U s i n g U c h i d a and T a g a m i ' s (1984) bodyweight to carapace length mode l , this equates to 0.123 k g and 1.33 k g , or 0.467 +185%/-74% k g per lobster. The dis tr ibut ion o f ghost f ishing mor ta l i ty as a percentage o f the reported landed catch is shown i n F igure 3.4.1. 116 Monte Carlo Figures Dist r ibut ions o f M o n t e C a r l o s imulat ions; x-axis denotes ghost f ishing as a percentage o f reported landings and y-ax is denotes frequency. The vert ic le l ine represents the result obtained from the ca lcula t ion o f midrange values. 500 500 400 300 200 100 0 0.00 0.05 0,10 0.15 Figure 3.4.1 - Hawaiian spiny lobster fishery 117 Chapter IV Discussion and Conclusion W h e n examin ing the results o f this study, several sources o f uncertainty must be acknowledged: i n the sampl ing , calculat ions, and the M o n t e Ca r lo analysis . S a m p l i n g U n c e r t a i n t y Estimates of fishing effort Estimates o f f i sh ing effort, i nc lud ing number o f gear i n use and trap loss rate, were m a i n l y obtained from surveys and l icense registrations. E a c h method has inherent uncertainties. L o w response rates o n surveys make it diff icul t to obtain a clear idea o f the situation i n the fishery as a who le , and responders m a y g ive deliberately mis l ead ing answers i f they feel that there is a po l i t i ca l agenda beh ind the survey. In case studies where f i shing effort is estimated f rom the number o f l icensed fishers or boats, it is assumed that a l l fishers are w o r k i n g fu l l t ime, w h i c h m a y be an overestimate since some fishers o n l y w o r k part t ime. Effectiveness of traps over time A l t h o u g h traps c o u l d be bur ied i n storms, corroded over t ime or destroyed b y woodborers (Breen 1990), B r e e n (1987) found that there was no difference between the catch rate o f traps at the beg inn ing and after one year. W i t h the except ion o f one case study, the operative span o f traps was estimated at less than 1 year i n the calculat ions, so it was assumed that the effectiveness o f the traps i n catching crustaceans was constant throughout its l ifespan. Behaviour and Anatomy Because different species o f crustaceans behave differently, they are often targetted b y different types o f traps, w h i c h affects the catch rate. True crabs m o v e i n a dis t inct ive s ideways scuttle w h i l e k i n g crabs are more spider- l ike i n their l o c o m o t i o n due to their long , spiny legs. Lobsters and sp iny lobsters c r awl s ideways and forwards w i t h their legs but s w i m backwards w i t h flaps o f their tails i f bursts o f speed are needed, al though spiny lobsters are more c lose ly related to crayf ish and do not have large front c laws . Case studies were sorted into four t axonomic groups for ease o f compar i son (Figure 3 .1) . In case studies where mor ta l i ty rate was a mis s ing parameter and had to be calculated as a mean o f other morta l i ty rates, values were appl ied f rom another fishery that targeted the same species i f avai lable. 118 One example o f the difference i n behaviour is whether caught ind iv idua l s attract or repel others o f the same species, a l though that cannot be necessari ly attributed to taxon-specif ic behaviour. In the Infraorder Brachyura , blue crabs are cannibal is t ic and are attracted to traps that already ho ld other b lue crabs ( G u i l l o r y 2001) . O n the other hand, behavioura l interactions between Dungeness crabs inside and outside traps result i n reduced catch ( M i l l e r 1979 ci ted i n B r e e n 1990). S i m i l a r l y , dead K i n g crabs i n a trap usua l ly repel other K i n g crabs ( H i g h 1985). L i v e lobsters i n traps attract addi t ional lobsters even after the bait is gone (Paul 1984). Measures of mortality A l t h o u g h the unit used for mor ta l i ty was individualstrap~ 1 year~ 1 , some studies were done over a per iod o f several weeks or months. In that case, the mor ta l i ty was extrapolated to a year, under the assumption that mor ta l i ty is constant throughout the year. T h i s m a y not be an accurate representation o f the situation, as B r e e n (1987) noted that the number o f crabs i n pots var ied throughout the year and that the number o f crabs found i n a trap is not a rel iable indicator o f the f i shing rate o f the trap because o f turnover and deaths. One o f his arguments for this case is that crabs might enter and then escape f rom a trap between observations; i n this thesis, it was assumed that the ingress and egress rates were equal, putt ing the number o f captured ind iv idua l s at equ i l ib r ium. O n l y direct mor ta l i ty is taken into account i n this thesis; it should be noted that the indirect mor ta l i ty or decreased fitness o f crustaceans as a result o f wounds obtained f rom escaping a trap was not examined. C a l c u l a t i o n U n c e r t a i n t y There are numerous potential sources for error i n the equations and parameters used i n the ca lcula t ion o f ghost f i sh ing morta l i ty . The linear nature o f the calculat ions means that a change i n one parameter d i rec t ly affects a l l der ived parameters. H a l v i n g the operating l i fespan o f a trap, for example , results i n h a l f the ghost f i sh ing mortal i ty . Other than Breen ' s (1987) study o f the B C Dungeness crab fishery, no case study had a l l the required parameters. A s a result, values from different articles and studies for v a r y i n g years and fisheries were put together to form one case study. The closest matches to the fisheries in quest ion were used, but the values used m a y not reflect the si tuation i n the fisheries because they m a y be f rom different years, different locations, or a smal ler geographic area than the one dealt w i t h i n the case study. In some references, authors gave values but d id not specify the loca t ion or year to w h i c h the values appl ied; in some cases, they d id not even supply units. In such cases, 119 guesses had to be made based on context and the publ ica t ion dates o f the references. In other cases, values were g iven i n units that were not appropriate for the case study calculat ions, so a corresponding number had to be inferred. F o r example, Par r i sh and K a z a m a (1992) expressed the mor ta l i ty rate o f ghost traps as 0.22 that o f w o r k i n g traps rather than as ind iv idua ls caught per trap per year, so the value was adjusted us ing the reported landings and the number o f traps i n use. In most case studies, one mean species weight was used for a l l case studies i n v o l v i n g the same species. T h i s is not an accurate representation, as mean weights o f crustaceans are different i n each fishery, depending on h o w heav i ly they are fished. M o n t e C a r l o U n c e r t a i n t y The quantity o f parameters avai lable for a case study does not affect the leve l o f ghost f i sh ing act ivi ty; it does, however , affect the range o f poss ible results. A t 10,000 runs, the shape o f the dis t r ibut ion can be seen clear ly , but i n many case studies, the range o f possible values is large due to the uncertainty i n the parameters. Parameters w i t h uncertainty were assumed to have symmet r ica l tr iangular distr ibutions, but i n some cases, the distr ibutions m a y not have been symmetr ica l . In a study o f degradation t ime o f traps i n A l a s k a , the mean t ime for degradation was not the midrange ( K i m k e r 1990). C o n c l u s i o n Despi te the potential sources o f error, a number o f conclus ions can be drawn f rom this study. M y estimates suggest that ghostf ishing as a percentage o f reported landings ranges f rom 0 .002% to 2 1 % w i t h 19 o f 44 case studies b e l o w 1%. The mean o f a l l estimates is 3.8%. The highest inc idence o f ghost f i sh ing occurs i n the 1973 At l an t i c inshore lobster fishery, where a combina t ion o f h i g h f ishing effort and h i g h rate o f loss take the ghost f ishing morta l i ty to 21 .4%. F igure 4.1 shows the frequency o f percentage classes result ing f rom the case studies. 120 35 0-5 540 10-15 15-20 20-25 Percentage Class Figure 4.1: Frequency chart of ghost fishing as a percentage reported landings Table 3 compares estimates o f ghost f i sh ing act iv i ty f rom the four t axonomic groups. The lowest incidence o f ghost f i sh ing i n crabs occurs i n Ireland, where l o w f ishing effort leads to a lower amount o f traps be ing lost. T h e highest incidence occurs i n M i s s i s s i p i , where ghost f i sh ing is h i g h w i t h respect to reported landings. Ghos t f i sh ing act iv i ty i n k i n g crabs is lower i n general due to legis la t ion i n A l a s k a requir ing the use o f 30 thread cotton twine o n traps. Th i s is not the case w i t h lobster fisheries, where the lowest occurence o f ghost f i sh ing relative to reported landings, i n N e w Eng land , comes about o n l y because the reported landings were very h igh . A s stated above, the h i g h ghost f i sh ing act ivi ty , i n the 1973 At l an t i c lobster fishery, is caused b y h i g h effort and trap loss rate. A s there was o n l y one case study i n v o l v i n g sp iny lobsters, the average, lowest and highest ghost f i sh ing act iv i ty have the same value; the l o w estimate is due to a l o w number o f lost traps. Table 3: Comparison of estimates by taxonomic group A v e r a g e (%) M i n i m u m (%) M a x i m u m (%) C r a b s 4.0 0.01 20 K i n g c r abs 0.52 0.002 2.6 L o b s t e r s 6.4 1.2 21.4 S p i n y lobs ters 0.024 0.024 0.024 A l t h o u g h the relationships used to estimate ghost f i sh ing mor ta l i ty are linear, two input parameters have a clear impact on ghost f i sh ing morta l i ty : duration o f ghost f i sh ing per iod and total number o f gears used i n the fishery. O f the 9 case studies i n w h i c h traps had an effective l ife o f less than 0.417 years (the value used for case studies mi s s ing 121 durat ion o f ghost f i sh ing per iod) , 7 had results i n w h i c h ghost f ishing was less than 1% o f reported landings. The remain ing 2 case studies i n v o l v e d a large amount o f f i sh ing gear i n use or re la t ively l o w reported landings. W h e n a large number o f traps is used, a smal l difference i n the rate o f gear loss w o u l d make a large difference i n the total number o f traps lost - w h i c h i n turn affects ghost f i sh ing morta l i ty . Ghos t f i sh ing mor ta l i ty reduces avai lable b iomass for catch or spawning. The case studies above show that ghost f ishing can be an issue i n many fisheries, even those where biodegradable panels or twines are legislated. In other fisheries where the use o f these biodegradable devices are not legislated or enforced, ghost f ishing mor ta l i ty is higher. Some measures have already been taken to m i n i m i s e ghost f ishing. In some jur isd ic t ions , leg is la t ion requires the use o f proper ly marked buoys , escape gaps, and degradable plast ics or galvanic t imed releases. Federa l l aw i n the U n i t e d States mandates escape vents and a degradable panel that w o u l d a l l o w legal s ized lobsters and most other prey to escape ghost traps (Carr and Har r i s 1987). S o m e fisheries use 'habipots ' that lobsters can use as shelter and enter and leave unhindered ( S m o l o w i t z 1978). S i m i l a r l y , decreasing the number o f lost traps can reduce the damage that they cause; for example , assigning trap-free lanes i n areas o f heavy boat traffic prevents buoy lines f rom be ing tangled or cut. Escape vents and biodegradable ports al leviate the h igh mortal i ty f rom ghost gear. The most effective w a y to decrease the effect o f f i sh ing effort w o u l d be to reduce the number o f traps i n use. W h e n f ishing effort decreases, so does the number o f traps used. The fewer the traps that are used, the fewer are lost. E v e n i f the trap loss rate were l o w , a large number o f traps i n use w o u l d mean that a large number o f traps w o u l d be lost. R e d u c i n g f ishing effort, unfortunately, is diff icul t and expensive to enforce (Breen 1990), and escape ports ameliorate but do not e l iminate the p rob lem o f ghost f i sh ing (Stevens et al. 1993). Fishers m a y also resist legis la t ion requi r ing biodegradable devices o n traps, especial ly i f they feel that ghost f ishing is not an issue i n their part icular fishery. It is possible to demonstrate the existence o f ghost f i sh ing w i t h sidescan sonar to record the entry and morta l i ty o f crustaceans i n pots (Stevens et al. 2000), but i f ghost f i sh ing act ivi ty is l o w , po l i cy -makers m a y decide not to enact legis la t ion to prevent it. A n economic evaluat ion, compar ing the loss o f revenue f rom ghost f ishing ac t iv i ty and the cost o f us ing biodegradable devices and enforcing legis la t ion, is one method o f just i f icat ion. In this case, the methodology proposed i n this thesis can be adapted to show the economic value o f crustaceans lost to ghost f ishing. Wha t level o f ghost f i sh ing 122 morta l i ty is c o m m e r c i a l l y and po l i t i c a l l y acceptable becomes a management issue rather than a b i o l o g i c a l one. Ghost f i sh ing is one type o f i l l ega l , unregulated and unreported ( I U U ) f ishing that has o n l y recently been exp l i c i t ly studied i n m u c h depth by fisheries scientists. E v e n w i t h i n the realms o f I U U f ishing, re la t ively li t t le attention has been g iven to lost gear. B y in t roducing a s imple yet effective methodology, this thesis a l lows exis t ing informat ion to be consol idated and examined for trends and patterns despite the dearth o f studies on ghost f i sh ing act ivi ty . A study o f this scope is a starting point towards a better understanding o f the impact o f ghost f i sh ing , and hopeful ly the precursor o f others to come. 123 References A l a s k a Department o f F i s h and G a m e . 2004a. 1999 Pre l imina ry A l a s k a C o m m e r c i a l She l l f i sh Catches & Est imated E x v e s s e l Va lues . In: A l a s k a Department o f F i s h and G a m e . A l a s k a Department o f F i s h and G a m e . 2004b. A l a s k a C o m m e r c i a l Harvests o f K i n g , Tanner, and S n o w Crab , 1953-2000 . In: A l a s k a Department o f F i s h and G a m e . A l v e r s o n , D . L . 1997. G l o b a l assessment o f fisheries bycatch and discards: A summary overv iew. A m e r i c a n Fisher ies Socie ty S y m p o s i u m , Seattle, Wash ing ton , A m e r i c a n Fisher ies Socie ty . Arcemen t E . and V . G u i l l o r y . 1993. Ghos t F i s h i n g i n V en t ed and U n v e n t e d B l u e Crab Traps. The Proceedings o f the L o u i s i a n a A c a d e m y o f Sciences 56:1-7. Arcement , E . and V . G u i l l o r y . 1994. Ghos t F i s h i n g i n V e n t e d and U n v e n t e d B l u e Crab Traps. The Proceedings o f the L o u i s i a n a A c a d e m y o f Sciences 56:1-7. B i s h o p , J . M . 1983. Incidental capture o f d iamondback terrapin b y crab pots. Estuaries 6:426-430. B l o i s , S.d. 1992. The implementa t ion o f the G a l v a n i c T i m e Release M e c h a n i s m on Queen Crab Pots i n the G u l f o f St. L a w r e n c e : A case o f more responsible f ishing. In: M T S '92 , g loba l ocean partnership : proceedings: Wash ing ton Sheraton H o t e l , October 19-21. Wash ing ton , D . C . : M a r i n e Techno logy Socie ty ; 415. Breen , P . A . 1985. Ghost F i s h i n g b y Dungeness Crab Traps: A p re l imina ry report. Canad ian Manusc r ip t Repor t o f Fisher ies and A q u a t i c Sciences 1848: 51-55. Breen , P . A . 1987. M o r t a l i t y o f Dungeness Crabs Caused b y L o s t Traps i n the Fraser R i v e r Estuary, B r i t i s h C o l u m b i a . N o r t h A m e r i c a n Journal o f Fisher ies Management 7:429-435. Breen , P . A : 1990. A R e v i e w o f Ghost F i s h i n g by Traps and Gi l lne t s . In: Proceedings o f the Second International Conference on M a r i n e Debr i s , 2-7 A p r i l 1989, H o n o l u l u , H a w a i i , N O A A - T M - N M F S - S W F S C - 1 5 4 ed ( R . S . Shomura and M . L . Godfrey , eds). H o n o l u l u : N O A A ; 571-599. B r o w n , C . G . 1979. N o r f o l k Crab Investigations 1969-1973. In. Suffo lk : M i n i s t r y o f Agr i cu l tu re , Fisher ies and F o o d , Fisher ies Laboratory , Lowestof t , Suffolk; 11. 124 B r o w n , S. K . et al. 1998. E c o l o g i c a l Effects o f F i s h i n g . N O A A ' s State o f the Coast Report . Na t iona l Oceanic and Atmospher i c Admin i s t r a t i on , S i l v e r Spr ing , M D . B u l l i m o r e , B . A . , et al. 2001 . A study o f catches i n a fleet o f "ghos t - f i sh ing" pots. F i she ry B u l l e t i n 99:247-253. Carr , H . A . and J . Har r i s . 1997. Ghos t -F i sh ing Gear : H a v e F i s h i n g Pract ices D u r i n g the Past F e w Yea r s Reduced the Impact? In: M a r i n e Debr i s : Sources, impacts, and solutions ( J . M . C o e and D . B . Rogers , eds). N e w Y o r k C i t y : Sp r inge r -Vr l ag ; 141-151. C h o p i n , F . et al. 1997. Sources o f Unaccoun ted M o r t a l i t y i n F i s h Capture Technologies . Pages 149-155 in D . A . H a n c o c k , editor. D e v e l o p i n g and sustaining w o r l d fisheries resources: the state o f science and management : 2nd W o r l d Fisher ies Congress . , C o m m o n w e a l t h S c i e n t i f i c & Industrial Research Organisat ion, Aus t r a l i a , C o l l i n g w o o d , V i c t o r i a . Center for Env i ronmen ta l Educa t ion . 1987. Plast ics i n the ocean: M o r e than a litter p rob lem. Center for Env i ronmen ta l Educa t ion , 1725 DeSa le s St., Wash ing ton , D . C , 128 p. Dah l s t rom, W . A . 1975. S u m m a r y o f informat ion o n recovered lost traps. C a l i f o r n i a Department o f F i s h and G a m e , M a r i n e Research R e g i o n , procedural report. 20 p . D e M a r t i n i , E . E . , G . T . D i N a r d o , and H . A . W i l l i a m s . 2003. T e m p o r a l changes i n popula t ion density, fecundity, and egg size o f the H a w a i i a n spiny lobster (Panulirus marginatus) at N e c k e r B a n k , Nor thwestern H a w a i i a n Islands. F i she ry B u l l e t i n 101(1):22-31. D e m o r y , D . 1971. A b a n d o n e d crab pots near C a n n o n B e a c h , Oregon . Fisher ies C o m m i s s i o n . Oregon She l l f i sh Investigation Information Repor t 70-6. 5 p. D o w , R . L . 1961. Some factors in f luenc ing M a i n e lobster landings. C o m m e r c i a l Fisher ies R e v i e w . 23(9): 1-11. Ebert , T . A . and R . F . F o r d . 1986. Popu la t ion ecology and fishery potential o f the spiny lobster Panulirus penicillatus at Enewetak A t o l l , M a r s h a l l Islands. B u l l e t i n o f M a r i n e Science 38:56-67. 125 Ef ron , B . and R . T ibsh i r an i . 1986. Bootstrap methods for standard errors, confidence intervals and other measures o f statistical accuracy. Statist ical Science 1(1): 54-77. F a h y E . 2001. The Magharees spider crab Maja Squinado f ishery i n 2000. D u b l i n , Ireland: M a r i n e Institute; 23 . Fahy , E . et al. 2002. T h e inshore pot fishery for b r o w n crab {Cancer pagurus) l anding into south east Ireland: estimate o f y i e l d and assessment o f status. A bbo t s tow n , D u b l i n : M a r i n e Fisher ies Services D i v i s i o n , M a r i n e Institute; 26. Fogar ty , M . J . and D . V . D . B o r d e n . 1980. Effects o f trap vent ing on gear select ivi ty i n the inshore R h o d e Is land A m e r i c a n lobster, Homarus americanus, fishery. F i she ry B u l l e t i n . 77:925-933. Garstang, W . 1903. The impover ishment o f the sea: A cr i t i ca l summary o f the experimental and statistical evidence bear ing upon the al leged deplet ion o f the t rawl ing grounds. Journal o f the M a r i n e B i o l o g i c a l A s s o c i a t i o n o f the U n i t e d K i n g d o m . 6:1-69. G i b s o n , F . A . 1973. C a t c h and effort i n the I r ish lobster fishery dur ing 1971. Department o f Agr i cu l tu re and Fisher ies i n Ireland F i she ry Leaflet N o . 52. l i p . God0y, H . et al. 2003. Unaccoun ted mortal i ty o f red k i n g crab (ParaUthodes camtschaticus) i n deliberately lost pots o f f Nor the rn N o r w a y . Fisheries Research 64(2-3): 171-177. Granath, K . L . 2002. F i she ry Information Packet for the B r i s t o l B a y R e d K i n g Crab Fishery , 2002. In. K o d i a k , A l a s k a : A l a s k a Department o f F i s h and G a m e ; 15. G r i f f i n , K . L . and M . L . W a r d . 1992. A n n u a l management report for the shel l f ish fisheries o f the B e r i n g Sea area, 1992. In: A n n u a l Management Report for the Shel l f i sh fisheries o f the westward region, 1991. A l a s k a Department o f F i s h and Game . D i v i s i o n o f C o m m e r c i a l Fisheries , R e g i o n a l Information Report 4 K 9 2 - 9 , K o d i a k . p. 151-197. G u i l l o r y , V . 1993. Ghos t F i s h i n g by B l u e Crab Traps. N o r t h A m e r i c a n Journal o f Fisheries Management 13:459-466. G u i l l o r y , V . 1999. A R e v i e w o f Incidental F i s h i n g Mor ta l i t i e s o f B l u e Crabs. In: B l u e Crab M o r t a l i t y S y m p o s i u m : G u l f States M a r i n e Fisher ies C o m m i s s i o n ; 28-41 . 126 Guillory, V., M. Bourgeois, P. Prejean, J. Burdon and J. Merrell. 1994. A biological and fisheries profile of the blue crab, Callinectes sapidus. Louisiana Department of Wildlife and Fisheries, Fisheries Management. Plan Ser. Part 1. Guillory, V., A. McMillen-Jackson, L. Hartman, H. Perry, T. Floyd, T. Wagner and G. Graham. 2001. Blue Crab Derelict Traps and Trap Removal Programs. Ocean Springs, MI: Gulf States Marne Fisheries Commission. Guillory, V. and J. Merrell. A survey of 1996 Louisiana commercial blue crab fishermen. Louisiana Department of Wildlife and Fisheries, Unpublished Report. Guillory, V., P. Prejean, M. Bourgeois, J. Burdon, and J. Merrell. 1996. A biological and fisheries profile of the blue crab, Callinectes sapidus. In: Louisiana Department of Wildlife and Fisheries, Fisheries Management Planning Service, No. 8, Pt. 1. Hammerschmidt, P., T. Wagner, and G. Lewis. 1998. Status and trends in the Texas (Callinectes sapidus) fishery. Journal of Shellfish Research. 17(2):405-412. Hartman, Leslie. Personal communication to Guillory et al. Alabama Department of Conservation and Natural Resources, Marine Resource Division, P.O. Box 189, Dauphin Island, Alabama. Heath, S. 1998. The Alabama blue crab fishery: historical trends, status, management, and future. Journal of Shellfish Research, 17(2): 435-439. High, W.L. 1985. Some consequences of lost fishing gear. In: Workshop on the Fate and Impact of Marine Debris (R.S. Shomura and H.O. Yoshida, eds). Honolulu, Hawaii: NOAA; 430-437. High, W.L. and D.D Worlund, 1979. Escape of king crab Paralithodes camtschatica from derelict pots. In: NOAA; 11. Industry Canada, 2004. Snow crab. In: Industry Canada. INFOFISH.1995. Article 13: Elliptical Traps. INFOFISH. 10. March-May 1995. Kelleher, K. Discards in the world's marine fisheries: an update. FAO Fisheries Technical Paper No. 470. 134pp. ftp://ftp.fao.org/FI/DOCUMENT/t470_advanced/t470_advanced.pdf 127 K e n n e d y , R . 1986. The Dungeness crab o f N o r t h A m e r i c a . Safish (South Aus t r a l i an Department o f Fisheries) 9:35. K i m k e r , A . 1990. Biodegradable twine report to the A l a s k a B o a r d o f Fisher ies . In. Anchorage , A K : A l a s k a Department o f F i s h and G a m e ; 10. K o d i a k Chamber o f Commerce . 2001 . K o d i a k R e g i o n Comprehens ive E c o n o m i c Deve lopment Strategy Augus t 2001 . In. K o d i a k , A l a s k a : A l a s k a Department o f F i s h and Game ; 65. Krouse , J .S. 1989. Performance and select ivi ty o f trap fisheries for crustaceans. In J .F . C a d d y (ed.). M a r i n e Invertebrate Fisher ies : The i r assessment and management, p. 307-325. N e w Y o r k : W i l e y . K r u s e , G . H . and A . K i m k e r , 1993. Degradable Escape M e c h a n i s m s for Po t Gear : A summary report to the A l a s k a B o a r d o f Fisher ies . In. Juneau, A l a s k a : A l a s k a Department o f F i s h and G a m e ; 24. Landgraf , K . C . et al. 1990. A n n u a l Repor t o f the 1989 western Pac i f i c lobster fishery. In: N M F S Southwest F i s h . Science Center.; 30 p. M a l l e t , P . Y . et al. 1998. A rev iew o f catch, f i sh ing effort and b i o l o g i c a l trends for the 1987 southwestern G u l f o f St. L a w r e n c e snow crab, Chionoecetes o p i l i o fishery. In: C A F S A C (Canadian A t l an t i c Fisher ies Scient i f ic A d v i s o r y Commi t t ee ); 39. M c M i l l a n - J a c k s o n , A n n e . Personal communica t i on to G u i l l o r y et al. 2001 . F l o r i d a F i s h and W i l d l i f e Conserva t ion C o m m i s s i o n , 100 E igh t A v e n u e S E , St. Petersburg, F l o r i d a . Merce r , L . P . 2002. M a i n e Department o f M a r i n e Resources Bureau o f Resource Management A n n u a l Report . In: M a i n e Department o f M a r i n e Resources Bureau o f Resource Management . M i l l e r , R J . 1977. Resource under t i l iza t ion i n a spider crab industry. Fisher ies (Bethesda) 2:9-13. M i l l e r , R . J . 1979. Saturation o f crab traps: Red u ced entry and escapement. Journal o f Conserva t ion . 38:338-345. 128 M u i r , W . D . et al. 1984. Escape o f captured Dungeness crabs f rom c o m m e r c i a l crab pots i n the C o l u m b i a R i v e r estuary. N o r t h A m e r i c a n Journal o f Fisher ies Management 4:552-555. M u r p h y , M . D . et al. 2001 . A stock assessment for blue crab, Callinectes sapidus, i n F l o r i d a waters. In. St. Petersburg, F l o r i d a : F l o r i d a F i s h and W i l d l i f e Conserva t ion C o m m i s s i o n ; 56. Na t iona l M a r i n e Fisheries Service . U n p u b l i s h e d data. Interview sheets o n file at the W o o d s H o l e Labora tory , Northeast Fisher ies Center, N a t i o n a l M a r i n e Fisheries Service . N O A A . W o o d s H o l e , M A . 02543. N a t i o n a l M a r i n e Fisheries Service . 2004. N M F S C o m m e r c i a l F i she ry L a n d i n g s Data . In: N a t i o n a l M a r i n e Fisher ies Service . N i p p e s , W i l l i a m . A l a s k a Department o f F i s h and G a m e . Personal communica t i on to K r u s e and K i m k e r 1993. K o d i a k , A l a s k a . Nor thup , T . 1978. Deve lopment o f management informat ion for coastal Dungeness crab fishery. Project C o m p l e t i o n Report , Project N o . 1-114-R. Pac i f i c M a r i n e Fisher ies C o m m i s s i o n . 1978. Dungeness crab project o f the state-federal fisheries management program. In. Por t land, Oregon: Pac i f i c M a r i n e Fisher ies C o m m i s s i o n . Par r i sh , F . A . and T . K . K a z a m a . 1992. Eva lua t i on o f ghost f i sh ing i n the H a w a i i a n lobster fishery. F i shery B u l l e t i n 90:720-725. P a u l , J . M . 1984. Investigations into escape vent effectiveness and ghost fishing i n captive populat ions o f the spiny lobster, Panulirus marginatus. Second S y m p o s i u m on Resource Investigations i n the Nor thwestern H a w a i i a n Islands, U n i v e r s i t y o f H a w a i i . P a u l , J . M . et al. 1994. Compensa tory feeding capaci t iy o f two B r a c h y u r a n crabs, Tanner and Dungeness, after starvation periods l ike those encountered i n pots. F i shery B u l l e t i n 90:720-725. Prudden , T . M . 1962. A b o u t lobsters. Freeport , M a i n : B o n d W h e e l w r i g h t C o . R a w l i n g s , N . D . , D . P . T o l l e , & A . J Barrett . 2004. MEROPS: the peptidase database. Nucleic Acids Res. 32 Database issue, D 1 6 0 - D 1 6 4 . http://merops.sanger.ac.uk/ 129 Reeves , J . E . and B J . Turnock . 1999. A l a s k a She l l f i sh Fisheries . In: O u r L i v i n g Oceans: Report on the status o f U . S . l i v i n g marine resources, 1999: U . S . Department o f Commerce . Roberts , K J . and M . E . T h o m p s o n . 1982. E c o n o m i c elements o f commerc i a l crabbing i n L a k e Pontchartrain and L a k e Borgne . L o u i s i a n a State U n ive r s i t y , Sea Grant Pub l ica t ion L S T - T L - 8 2 - 0 0 1 . B a t o n Rouge . She ldon , W . N . and R . L . D o w , 1975. Trap contributions to losses i n the A m e r i c a n lobster fishery. F i she ry B u l l e t i n 73:449-451. S h i v e l y , J . D . 1997. Degradab i l i ty o f natural materials used to attach escapement panels to blue crab traps i n Texas . Texas Parks and W i l d l i f e Department, F i n a l Repor t . S K Project N A 6 7 F D 0 0 3 4 . S m o l o w i t z , R . J . 1978a. Trap D e s i g n and Ghost F i s h i n g : A n ove rv i ew . M a r i n e Fisher ies R e v i e w 40. S m o l o w i t z , R . J . 1978b. Trap D e s i g n and Ghost F i s h i n g : D i s c u s s i o n . M a r i n e Fisheries R e v i e w 40. Stevens, B . G . 1996. C rab B y c a t c h i n Pot Fisheries: Causes and Solut ions . In: S o l v i n g byca tch : considerations for today and tomorrow, September 25-27, 1995, Seattle, Wash ing ton . (Program A S G C , ed). Fai rbanks , A l a s k a : U n i v e r s i t y o f A l a s k a ; 151-158. Stevens, B . G . et al. 1993. Underwate r observations on behavior o f K i n g Crabs escaping f rom crab pots. In: A F S C ( A l a s k a Fisheries Science Center) . Stevens, B . G . et al. 2000. Ghos t f i sh ing b y Tanner crab (Chionoecetes bairdi) pots o f f K o d i a k , A l a s k a : pot density and catch per trap as deermined f rom sidescan sonar and pot recovery data. F i s h e r y B u l l e t i n 98:389-399. Tegelberg , H . C . 1974. Dungeness crab study, annual report, J u l y 1, 1973 to June 30, 1974. In: U . S . Department o f Commerce , Na t iona l M a r i n e Fisher ies Service , Wash ing ton . Texas Parks and W i l d l i f e Department. U n p u b l i s h e d data. 4200 S m i t h S c h o o l R o a d , A u s t i n , Texas. 130 U c h i d a , R . N . , and D . T . Tagami . 1984. B i o l o g y , dis tr ibut ion, popula t ion structure, and pre-exploitat ion abundance o f spiny lobster, Panulirus marginatus (Quoy and G a i m a r d 1825), i n the Northwestern H a w a i i a n Islands. In: Proceedings o f the second sympos ium on status o f resource investigations i n the Northwestern H a w a i i a n Islands (R. W . G r i g g and K . Y . Tanoue, eds.), p. 157-198. Sea Grant Misce l l aneous Report U N I H I - S E A G R A N T - M R - 8 4 - 0 1 . Sea Grant Program, Un ive r s i t y o f H a w a i i , H o n o l u l u , H a w a i i . V ienneau , R . and M . M o r i y a s u . 1994. Study o f the impact o f ghost f ishing on snow crab, Chionoecetes o p i l i o , by convent ional con ica l traps. Canad ian Techn ica l Report o f Fisheries and Aqua t i c Sciences: i -9. Whi take r , J . D . et al. 1998. A rev iew o f the fishery and b i o l o g y o f the blue crab, Callinectes sapidus, i n South Caro l ina . Journal o f Shel l f i sh Research 17:459-463. 131 Appendix 1: Visual Basic Code for Monte Carlo Simulations Parameter (Par) list: Par(l): Total gear Par(2): Annual trap loss rate Par(3): Gear lost per year Par(4): Percentage of lost traps ghost fishing Par(5): Number of traps ghost fishing Par(6): Length of ghost fishing period Par(7): Number of traps ghost fishing per year Par(8): Mortality rate per trap Par(9): Annual ghost fishing mortality Par(lO): Average weight of species Par(l 1): Weight of annual ghost fishing mortality Par(12): Reported landings PerGhostCalc: Ghost fishing as a percentage of reported landings ParL() is the lower limit in the range of the parameter ParU() is the upper limit in the range of the parameter ParM() is the mean in the range of the parameter — Begin Code — Sub CaseStudyAnalysis() Dim i As Integer, j As Integer, CS As Integer, trials As Integer Dim PerGhost() As Double Dim ParL(12) As Single, ParU(12) As Single, ParM(12) As Single Dim ColCount As Integer, pcount As Integer Dim PerGhostCalc As Double Dim xx As Single, Ntrials As Integer, NCases Worksheets("Case Studies tri values"). Activate Ntrials = 10000: NCases = 60 ReDim PerGhost(NCases, Ntrials) For CS - 1 To NCases ColCount = 8 pcount = 0 Do pcount = pcount + 1 ParM(pcount) = Cells(CS + 2, ColCount) ParL(pcount) = Cells(CS + 2, ColCount + 1) 132 ParU(pcount ) = C e l l s ( C S + 2, C o l C o u n t + 2) C o l C o u n t = C o l C o u n t + 3 L o o p U n t i l pcount = 1 2 F o r tr ial = 1 T o Nt r ia l s x x = 0 I f P a r M ( l ) > 0 T h e n I f P a r U ( l ) = 0 T h e n I f P a r M ( 2 ) = 0 T h e n x x = P a r M ( 3 ) E l s e x x = P a r M ( l ) E n d I f E l se x x = ( ( rantr iangle((ParL(l) * P a r M ( l ) ) , ( P a r U ( l ) * P a r M ( l ) ) ) ) + P a r M ( l ) ) E n d I f I f Pa rU(2) = 0 T h e n x x = x x * P a r M ( 2 ) E l se x x = x x * ((rantriangle((ParL(2) * P a r M ( 2 ) ) , (ParU(2) * ParM(2) ) ) ) + P a r M ( 2 ) ) E n d I f E l s e x x = P a r M ( 3 ) E n d I f I f P a r L ( 3 ) > 0 A n d ParU(3) = 0 Then x x = P a r M ( 3 ) I f P a r L ( 3 ) > 0 A n d ParU(3) > 0 T h e n x x = ((rantriangle((ParL(3) * P a r M ( 3 ) ) , (ParU(3) * ParM(3)) ) ) + Pa rM(3) ) I f P a r U ( 4 ) = 0 T h e n x x = x x * P a r M ( 4 ) E l s e x x = x x * ((rantriangle((ParL(4) * P a r M ( 4 ) ) , (ParU(4) * ParM(4) ) ) ) + Pa rM(4) ) E n d I f I f P a r U ( 6 ) = 0 T h e n x x = x x * P a r M ( 6 ) E l se x x = x x * ((rantriangle((ParL(6) * P a r M ( 6 ) ) , (ParU(6) * ParM(6) ) ) ) + Pa rM(6) ) E n d I f 133 I f ParU(8) = 0 Then x x = x x * ParM(8) E l s e x x = x x * ((rantriangle((ParL(8) * ParM(8)), (ParU(8) * ParM(8)))) + ParM(8)) E n d I f I f P a r U ( 1 0 ) = 0 T h e n x x = x x * P a r M ( 1 0 ) E l s e x x = x x * ((rantriangle((ParL(10) * P a r M ( l O ) ) , ( P a r U ( l O ) * P a r M ( l O ) ) ) ) + P a r M ( l O ) ) E n d I f I f P a r U ( 1 2 ) = 0 T h e n P e r G h o s t C a l c = x x / P a r M ( 1 2 ) E l s e P e r G h o s t C a l c = x x / ((rantriangle((ParL(12) * Pa rM(12) ) , (ParU(12) * Pa rM(12) ) ) ) + Pa rM(12 ) ) E n d I f P e r G h o s t ( C S , trial) = P e r G h o s t C a l c N e x t t r ia l N e x t C S Worksheets ("Temp") . Ac t iva te F o r C S = 1 T o N C a s e s F o r i = 1 T o Nt r i a l s C e l l s ( i , C S ) = P e r G h o s t ( C S , i) N e x t i N e x t C S E n d Sub F u n c t i o n ran t r i ang le (Er rLow A s Var ian t , E r r H i A s Var iant ) A s Single D i m U l A s S ing le D i m U 2 A s S ing le D i m U 3 A s S ing le 134 D i m U 4 A s Single D i m T l A s S ing le D i m T 2 A s S ing le D i m F A s S ing le D i m P A s Single E r r L o w = A b s ( E r r L o w ) E r r H i = A b s ( E r r H i ) U l = E r r L o w * R n d U 2 = E r r L o w * R n d T I = - A b s f U l - U 2 ) U 3 = E r r H i * R n d U 4 = E r r H i * R n d T 2 = A b s ( U 3 - U 4 ) F = R n d P = E r r L o w / ( E r r L o w + E r r H i ) I f F < P T h e n rantriangle = T I E l se rantriangle = T 2 E n d I f E n d Func t ion 135 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0074876/manifest

Comment

Related Items