Pollachius Virens L) Escaping Through Trawl Meshes

...

ICES PAPER C.M. 19921B:29 Fish Capt. Cttee Ref.G

SURVIVAL OF SAITHE (POLLACHIUS VIRENS L) ESCAPING THROUGH TRAWL MESHES

Jan Arge Jacobsen1, Bjarti Thomsen1 and Bjomar Isaksen2

1 Fiskirannsoknarstovan Noarun, FR-lOO Torshavn Faroe lslands

2 Havforskningsinstituttet Fangstseksjonen, N-5024 Bergen Norway

ABSTRACT

A pilot study to estimate the survival of saithe escaping from a cod-end with 145 mm diamond meshes was done in April-May 1992 north of the Faroes. Escaped fish were collected in fine meshed net cages (2x2x5m aluminium frame) mounted aft on a cod-end cover. After one,hour trawling below 150m depth the cages were released by means ofan acoustic release system and slowly hauled up to 40m below sea surface for UTV observations (6-7 days). The net cages were drifting freely with the current in the area north of the Faroes and located by means of radio tracked buoys. Preliminary results indicate that saithe can withstand almost the same sorting as cod with low mortality. However, more experimental work is needed to draw more firm conclusions on survival rates ofsaithe escaping from a cod-end. 2

. , INTRODUCTION

The basic philosophy in fishery management. by mesh size regulation of trawlfisheries is to let fish under a certain minimuffi size escape, .thereby increasing the potential futUre yield of the fisheries. A necessary requirement is t1ult the escaiJing fish survive. But how badly daIDaged are the fish that manage to escape from a cod-end? Do they survive? These questionS are impcirtlnt, arid attempts to answer them \....ere doneby among others Mairie and Sangster (1988), Zafemian arid Serebrov (1989), and recently Soldal et al. (1991). The first results indicated a. high mortalitY ofyoung haddock in the North Sea (Maine and Sangster 1988, 1989). The SUMVal estimätes of haddock were. adjusted upwardS in later reports due to more. refined techiUques and bettel' methodology (Maine and Sangster 1990, 1991, arid Soldal et al. 1991). The species stUdied so far are haddock, cod, whiting, and to a lesserexterii saithe. These snidies were done in the North Sea, and in Northelll Norway. Baltic herrmg has also been studied reeently (Suuronen 1991).

The commercial trawl fishery at Faroes is inainiy baSed onfresh-jish stern trawlers, traditionally aimed at cod. Howevcr, a dciStic declirie in the cod fisheiy in later yeacs has initiatoo a shift of prmcipal target species to\....ardS saithe. This change led to a conversion of most of the trawler fleet to fish on deeper \\'aier as pair trim'lers to catch saithe. AS the mesh size in the demerSal fisheries in the Faroes EEZ is 145 min, substariiial amotÜtt of young saithe is exiJected to be • selected out. It is therefore necessarY to krlow if saithe escaping from cod-ends do survive.. The work done so far can not be transferred to the Faroese fishery directly, as the studies have been done iri sha110w \\'ater less than 60 m depth, while saithe is caughi at depthS below 150 ~ also the mesh size is different to those studied and lastly the survival data available on saithe is rather liriUted (Soldal et al. 1991). The aim ofthe work presented in this paper ,....<15 thus to ehiddate thci survival rates ofsaithe escapmg from a cod-end.

, . MATERIAL AND METIIODS

• .I • The experimentS were conducted iri April-May. 1992 man area ncil-th of the Faroes (Figure 1) onboard the chartered vesseI Mrr Llggjas (2hn, 370Hp) and the research vessel Afagnus Heinason (45m, 1800Hp). Scientific personnel from Denmark, Nom'aY and Faroes pariicipated in the sui-Yey.

As the flshing areaS for saithe are located relatively far offthe coast at depths below 150ni, it \\'aS considered lethal tothc fish to be towed in cages 15-20 nm to an inshore area for closer obserVa tion, aS used iri the Norwegian method. The possibility to anchor the cages 3.t bottom were consid ered. However, the current in the shelf-edge area is strong and this method provided DO possibilitY to inonitor the fish and the cages during the experimental period. Therefore the cages \\'ere hauled up to 40m for observation, where they were left drifting freely with current.

The sun-ey started on 22 April 1992, and the first week \\'aS used to fmd an ar~ north or north \~iest ofthe Faroes ,\ith suitable concentrntions of saithe of a \\ide size range, and at the same tinie not too far from the isles. The residual flow north ofthe Faroes is going east-wardS (Hansen 1979, 1991). It '\'aS thcrefore natUral to aim at an area. north-west ofthe isles, so the relciaSoo fish cages would drift in a fairly kßo\\n direction (Figure I).

The method usoo iri the present investigation can iri short termS be considered to be :in offshore version ofthe Norwegian method (Soldat et al.. 1991) adapted to deeper water. Qnly the differ ences from the Norwegiari method \\i11 be described in detail here. Trawling \\'aS perfomied below 150m depth \\ith a commercial bottom tm,,"l (Stjomutrol) \\ith 145 mm cod-end. The experimental set-up duriIig trawling is shO\\n i Figure 2. A smaU meshed cod-erid cover kept open \\ith aluffiinium rings (2m 12') was attached to the extension-piece on the tm\vi imd the escaping

\ 3

fish were collected in a fme meshed fish cage (2,ax5m ahilninium frame) mounted an on the cOd~ end cover (Figure 3).

After 1 hour trawlmg the cage waS released from the cod-end cover by means of a remoteiy controlIed acoustic release system arid then taken over by the small assistant vessel, which begari to haul the cage up to 40 m depth. This process was considered to be quite crucial in the \\'hole expenment. We tried to haul the cage very slowly, and only redueing the atniosphenc pressure in Steps by 1/4 eaeh 80-100 minute according to a decompression table (see T)1ler and Bhixter 1973, Efanov et a/. 1986). At the observation depth, usually 8-16 hours after release, 2,,4 plastic floats (11" 0) and two buoys (24" 0).were attached to the line. A marker buoy ,~..ith light arid ä rnwo buoy were also attached, to be able to locate the different cages diifting in the ärea by the assistant vessel (Figure 4). Tbe cages ' ....ere observoo daily by an UTV camera lowered do,..n the buoy line and the operator could observe the fish in the cage (Figure 5). All UTV observations were recorded on tape.

An attempt was Inade to analyse the scäie loss and skin damage ofiridividual fish as described in Maine arid Sangster (1988, 1989). But it soon became clear that the scale loss in most fishes was more severe on one side than on the other. This phenomenon ,,,'as meist common among the smallest fish arid especially for young haddock. Generally the scale loss on saithe ~,\'aS lo,v, but no • estiffiates were calculatcd. Tbe reason will bc dealt with in the discussion. Tbe survival of saithe escaping from a cod-erid was compared to a contro/ group; i.e. saithe collected mthe cage from an open trawl wherc the cod-end was removed. Tbe control cage and the cage ,..ith COd-end cseapecs were released the same day very elose together (we also tried to join the two cages) to secure that both cages experienced approximately the same environmental cciriditions during the observation periOd of6-7 daYs.

'\. , RESULTS

The weather conditioris were poor and we experieneed one severe gale in the middle ofthe surVey. AS the method is quite complex, this hampered the experiment significantly. Tbe first sucCessful rdease of an experimental cage eoribining adcqüate number of saithe was on the 28 April, one weck after sUrVey start. A control cage was also released later the same day (see IA and 2A in Figure 1). Several cages containing cod-end escapees arid control cages \vere released the following days. However, most of them bad to be excluded from the aßalysis. Some cages contairied DO saithe and others got ruptures in the cage netting. In three cases the cage ,,-as not released properly at bottom by the acoustic release s}'steni. Two cages came to sunace by themselves after release, mostly due to huge amount of small redfish (s. viviparus) whieh got positively buoyant. at a certain stage of the häuling proeess. Cage riumber 4 ,..-as successfully releaSed on the 1 May (see 4A* in Figure I), however, dunng the storm that day we lost the connectioD with the radio buoy, it ,..-as never reeovered. Four cages (no. 5,6,7 and 8) were' released during the seeond weck ofthe survey and the last cage was recovered on 10 May (see 8B in Figiire 1) by the assistaßt ,'esset.

Saithe. Due to the development and adjustIllent of the experimental technique arid bad weather, orily 1'....0 eages contained suitable number of saithe to calculate suiVival estimates (see eages 1 and 2 in Table 1). All saithe mthe control cage survived confmeinent äfter 6 days, v..hile only 2 out of 85 fish (2.4%) died in the cage containing cod-erid escapees (fable 1). Subsequent post mortern exaIniruition of the 1'....0 fishes revealed erupted swim bladder, and no signs of eXternat damage were evident in either fish. Some ofthc sürvivcd saithe had net marks around the snout region probable due to pushing its snout into the meshes in the net ' ...aU. GeneraUy no sigils of fungus development \\"as eVident on the fishes, even if scale 10ss and skin darriage could be

\ 4

observed. TIus might be due to the relative short observation period of6-7 d:iys. However, Mairie and Sangstel' (1991) reported beginning signs ofdetenoraiion due to sca1e loss only after 4. d:iys and subsequent death within the nexi two da}'s, and Soldal et al. (1991) repolted heavy infection on all damaged haddock arid death withiri the neX! 5 d:iys and praetieally no deathS after one ,,,eek in captivity.

The length distrÜ>ution of saithe escaped from the eoo-end ninged from 30-65em and the saithe retained in the ordinary cod-end ranged from 40-80em (Figure 6).

Other speeies. Of other speeies eaught "as rriainly redfish, and espeeially the small redfish (S. ViViparus). Nearly aU redfish died after 1-3 days in captivity, ",ith only 25% survival in the Control group and 0-5% survival in the experimental group (Table 2). The S\\im bladder was not erupted iti the dead redfish. The size ofthe few redfish that survivoo was notably higher thah the bulk ofthe dead fish, indicating size dependent sumval. However this has to be verified in futUre work. A few halibut were caught duririg thc sUrVey (size ranges from 40-75em), they all died . äfter, a couple ofdays in eaptivity. The same was true for speeies as ling, tusk, and monkfishes. The ling and tusk also bad erupted swim bladder. "- • Analysis of video recordings..The UTV observations of the floating cages were deine daily (Figlire 5), and sho\ved the saithe swimming ca1m1y around the coOflnement oftheir cages. Most round fishes swam calmly ill'ound, even ifthe cage moved continuously ,"ith the waves arid aS.the line up to the markerbuoy was heaving in the cage. However, halibut arid other bottorn dwelling fish as catfish, monkfish arid ling seemed to be more dishubed arid excited thari "normal" under this inoving "boltorn".

,< , DISCUSSION

Tbe method. The present work must be considered as a pilot study to mOdify and develop an experimental technique on sUMval of cod-erid escapees in the open sea arid at depths below 150m. The depth, the offshore loeation ofsaithe, arid the strong current along the shelfedge were all factorS that made the experiment ehallenging, but at the same time they limited the experinient in several ways. Tori few cages were released successfully containing adequaie nUmber ofsaithe. to draw any firin condusions on sumval rates. Also the fast drifting of the floating cages caSt \·,

The technical equipment, the trawl. the cod-end cover (kept open by 2m 0 hoops), and thecages (Figure 3) worked exceptionally weU duririg the survey. Video reeordings ,..ithin the cage during trawling showed deal' space between the cOd-end cover and the cOd-end due io the aluffiiniuin ririgs. No maskirig is considered to have occurred.

Survival. Tbe two dead saithe had both erupted s,..im bladders. No supel-flcial signs ofscale lass 01' skin dailiage ,~..ere evident that could have causoo their death. A possible death cause eould ru1ve been a to rapid decompression, \\·heri the cages ,~..ere hauled up to observation depth of40m. If depth sensors bad been on the cages, a more careful decompression regime. could ru1ve been managed. GeneraIly tlle seaIe loss '·..as low. and in the few cascs .where it could be observed, an uneven damage.on each side waS noted, indicating that exllausted fish might have laid back arid rubbmg one side against the small meshed netting in the haltom of the tage during trawling 5 iasting one 1 hour. Soldal et al. (i991) mentionoo this phenomenon in small haddock after onty 10 min. trawling. Therefore no scale loss eStimates \,'ere calculated.

Beside the possible death cause by the to rapid decompression of fish, one additiorial cause cf death among haIibut, Img arid catfish, could have been the contiIiuos movement ofthe cages \\ith the \,'aves, thereby strcssing the fish usuaIly adaptecl to rest on bottoln. Indeed the dead halibut had clear signs ofdaßuged blind side.

One unexpected resuit was the high mortality of redflsh, and particularly of small redflsh (s. Viviparus), whether it hiui escaped the cOd-end or not (Table 2). One plausible explanation lriight be that redflsh, usuaIly considered a deep water species, is. neutriUly buoyant at depth thereby experiencing greater problems during depression, while roundfish might be negative buoyant at sea bed and Ileutn111ybuo)'aIlt in midwater, aS suggeStCd by Arriold (1991), and therefore better can \"ithstärid the hauling process.

... I. ,", To conclude:The present study haS revealecl and clarifiecl some of the expenmental deficiency enoountered durlI1g this survey, the techriiqtie has been tested arid acIjusted to work offshore on deep waters, and a few results have emerged. The method is now considered ready to give reliable survival estiniates on saithe and other species in future work.

Norwegian investigätions on the moitaIitf rate offish escäping tlu-ough tra\"l meshes indicated no mortality for Cod and amoitality less than 10% for haddoek (Soldal et al. 1991). The däta on saithe \vere insufficient to give reliable estiniates of mortality, although they stated that the mortality rate was lo\ver thari for haddock and probably higher than for cOd. The preliminary results from our first expenment tndicate that saithe can \"ithstand almost the same sortiIig iIi a COd-end as Cod with low mortality. However due to data deficiency more experimental work is neecled to dni\V more finn conchisionS on survivaI ofsaithe escäpmg from a eOd-end.

ACKNOWLEDGEl\IENT

The \"hole experlI11ent went smoothly due to the net män l\fogens Aridersen (DK), the techniCal assistant Oddvar Chfuickshank (NO), thc crew onboard MIS Llggjas arid RN Afagnus Heinason. The crews were indeed very co-operative during the experiment and had the respon sibility to mount the 'cages before the survey started. We would also like to extend our sincere thänks to the Noi-di6 Minister Coundl (Nordisk Ministerräd) for direct financial support during this experiment.

REFERENCES

Aniold, G.P. 1991.. Tbe depth of neutral buo)'ancy in cod: Verticai nugration arid systematic variation in Target Strength. ICES FAST \VG. Bergen 1992: Working paper.

Branov. S.F., Ts\'etkov, V.I., Abrasumov, V.A. and Istomin. lG. 1986. On the decompression regime of fishes lift from depth to the sufrace. ICES CM 1986 $:14). [Mimeo.)

iIärisen, B. 1979. Residual flo\v and tempernture on the Farce Plateau dUring the first half cf lin8 in relation to cirCulation. ICES CM 1979 (C:18). [Mimeo.)

HanSen, B. 1991.. Satellite-tiacked drogue paths over Farce Bank and thc Farce -ICelarid Ridge. leES CM 1991 (C:25). [Mimco,) 6

Maine, J. and Sangster, G.I. 1988. A progress report on an investigation to assess the scale damage and survival ofyoung gadoid fish escaping from the cod-end ofa demersal trawl. Scottish Fisheries Working Paper No 3/88.

Maine, J. and Sangster, G.I. 1989. Techniques for assessing fish damage and survival after escape from trawl cod-ends. Scottish Fisheries Working Paper No 9/89.

Maine, J. and Sangster, G.I. 1990. An assessment of the scale damage to and survival rates of young gadoid fish escaping from cod-end ofa demersal trawl. Scottish Fisheries Research Report No 46. 1990.

Maine, J. and Sangster, G.I. 1991. Do fish escaping from cod-ends survive? Scottish Fisheries Working Paper No 18/91.

Soldat, A.V, lsaksen, B., Marteinsson, J.E. and Engäs, A. 1991. Scale damage and survival of cod and haddock escaping from a demersal trawl. ICES CM 1991 (B:44). [Mirneo.]

Suuronen, P. 1991. Tbe effects of a rigid grating on the selection and survival of Baltic herring - pre liminary results. ICES CM 1991 (B:171. [Mirneo.]

Tytler, P. and Blaxter, J.H.S. 1973. Adaptation by cod and saithe to pressure changes. Netherl. Joum. of Sea Res. 7: 31-45.

zaferman, M.L. and Serebrov, L.I. 1989. On fish injuring when escaping through the trawl mesh. ICES CM 1989 (B:181. [Mirneo.] 7

9° 63° r-----,.--...,------___,.--__--.,---...... 100 m Main direction of current

.--...... 200 m ' .... _---- 1----t----t--,4A*-f------500 m ,.,- ·········5-SA-3B ·-~ ..·-·-8A7A " , , ,, I ·•·•·•·•·...1B "'" 5B \ / f--~-"\...-···-...·.. 2B ' SB 1-+\--+_11"'---> 30' I 30' I ..•.•. ~I " 7B. 8B , ",;' ) ..-...... " ! ,...... 620 1---'+---+------'''+-'-.,....-:'--/4<- "...... , ,,' t ' ...... \ , \. , \ ", .. " .'. \

/ ; 30' \ ....-.... ,,,,,, 30' ,----' ... , ..~ , I ...- , I ", , '. I , , i.· .• •-' I , , t

9°

Figure 1. Survey area. The numbered markers labelIed A show release positions of the cages, while the numbers labelIed B show the recovery positions. As can be seen, cage number 1 drifted over 75 nrn eastward during 6 days! Cage number 4 was never recovered, and is probably drifting towards the northem Norwegian coast by now. 8

Markerbuoy

150-200 m bottom depth

Cod-end cover with alu. rings (hoops) Fish cag Trawl

Ordinary cod-end

Figure 2. Experimental set-up during demersal trawling with the cage attached to aft end ofcod end cover (see details in Figure 3).

Line up to marker buoy (400-600 m)

eod-end eover with alu. rings (2m 0)

Acoustie release system \ • Tbe cage is attached/ to end ofeod-end cover Ordinary eod-end (145 mm)

Figure 3. Details ofthe attachment ofthe fish cage to the small meshed cod-end cover and the remotely operated acoustic release system. Fish escaping through the cod-end meshes is caught in the fine meshed net cage (2x2x5m aluminium frame). 9

Cage with fJsh H:::::::;;;=::::::::i::::~ Sm

Figure 4. Fish cage drifting freely with the current at 40 m depth during the observation period (4-7 days).

Marker buoy with light

Figure 5. Daily observation ofthe fish in the cage bya remotely operated urv. •.

Cl) 25 .c :!: GJ Ordinary cod-end ca 20 tn 't- 0 15 L. Cl) oe 10 E ~ Z 5 0 30 35 40 45 50 55 60 65 70 75 80 e Length (ern)

Figure 6. Length distribution of saithe from all trawl hauls during the survey; the dark area is the fish escaped from the cod-end and \\ithheld in the cod-end cover or the cages. The light area is fish re~ined in the ordinary 145 nun cod-end.

Table 1. Mortality results for saithe during the observation period (6-7 days). Only cage 1 (mesh selection) and cage 2 (contral) contained suitable number ofsaithe to calculate survival estimates.

Experimental group Cage no. Obs. period No. ofsaithe No. ofdeaths Sun"ival (%) Mesh selection 1 6 85 2 97.6 Mesh selection 6 6 2 1 Mesh selection 7 7 2 0 Mesh selection 8 7 3 0 Control 2 6 15 0 100.0

Table 2. Mortality results for redfish (s. viviparos) during the observation period (6-7 days).

Experimental group Cageno. Obs. period No. ofsaithe No. ofdeaths Survival (%) Mesh selection 1 6 17 16 S.9 Mesh selection 6 6 900 896 0.4 Mesh selection 7 7 8S 8S 0.0 Control 2 6 24 18 25.0