Internatonal Councl for -ehe Exploraton of the Sea COrr'.parat ve F~s hng C ornm ttee No. 8 The Selectvty of Halbut Gll Nets by SQ Olsen & J. Tjemsland Introducton In 936 gll nets were ntroduced n the Norwegan halbut fsheres, and n a short tme a successful gll-net fshery for large pre-spavmng and spaymng; halbut developed n all major spawnng areas. Prel7'ously, the nets \fvere made of hemp, but n recent years a change to polyamde fbres has taken place. Untl noyv the mesh sze has been regulated by law not to be less than 20.88 cm between the knots (about ls!; mesh s.'ze), but L.'. Ju.."le -r..:ls year ths regulaton 'Vvas changed to prohbt the use of mesh szes less than about 8 t T'ne fshng s carred out n the,!\lnter and the nets are set n long fleets at the bottom n 240 to 300 fathoms depth. Materal ~'.d Methods Durng the years 957 to 960 expermental fshng 'Wth nets of dfferont mesh szes and dfferent materals 'Were conducted ;n January-Februar;:v n one part!;cular area n. northern Norway. The nets used 'Were hemp and nylon nets of about lsh mesh sze and nylon nets nth mesh szes of approxmately 9!! and 23!l. Measured under a tenson of 4 klos the LlGsh szes averaged 42.8 cm, 49.6 cm and 6.9 cm respectvely" for the three dfferent types of nylon nets. The nets 'Were spread n groups of" tvvo to fve nets of each -type throughout the fleet, and 'wth changng postons of th:.; groups durng the se2.son. The type of nets for each ndvdual fsh Was recorded and the total length YJas measured. In 959 and 960 grth measurements 'Were also taken. The data collected n 957 and 958 showed that observatons on methods of attacbment -VIlere needed and for the two last years ~ experments records of such observatons were provded. For ths reason only the materal collected n 959 and 960 has been used L.'. ths treatment. In analysng the data the methods of attachment were grouped n ) meshng by the operculum and pont of greatest grth, and 2) all other methods of attachqent (.oo meshed by the maxllae, attached by the teeth, entangled by the tal, ccmpletely embedded n the net etc.). Table below gv'es the numbers caugh"c as PGI' type of net and attacr~ent method. Table. Nurnbers of halbut caught n 959 and 96c; specf8d by type of net and method of s"ttachillent. -,----------------~---------------------~-----~---~--- ---,! -.JJumbers caught : I ---- ~--I---- -------------- -----~----, Type of net Attacbment l Attachment 2 T-~~mp ~------r---------~~------------t-----------2;------+- I Nylon 6!! I 45 44 Nylon 9 j 92 29 ~ylon ~3T! ~_ -.l 25.. Calculaton of selecton curv"es For the meshed fsh (attacbment group ) selecton curv'es for the "Gnree dfferent mesh szes of nylon nets were calculated followng the method descrbed by Holt (957)~n l.::-:~.. (~_. ~xp. - ( - ID) 2 / 0 2 'Where: and n l s the number of 'sh of length caught \n s the mean selecton 8ngth.
- 2 - Table 2 ( page 4,) gves the length dstrbuton and log. ratos TIhen adjusted for dfferences n effort (no. of nets fshed)~ and n Fgure the log. ratos are plotted aganst length and lnes ftted by the method of' least squares are dravm. From the lnes of best ft v'alues for K, Im and ()2 Vere computed~- 2 b K = <'l K 9 Am A = K. Be Bm and 2 er -, 2 a where K s the rato between mean se0;6cton length, and mesh sze, 8, and a and b are coeffcents of the equaton y = 0. + b descrbn~ the lne of best ft for the log. ratos. The followng values for K were found~- KBA ::: 3.325, KCA = 3.33 and KCB = 3.54 The ndces A, B and C refer to the mesh szes 6!!, 9 t Ehe art-hmetcr mean of these, K "" 3.204 gves: and 23!l respectv'ely. A~m = 36.96 cm, Blm = 58.72 cm, varance ()2 = 2886. The selecton curves establshed n ths way for meshng by the operculum or pont of' greatest grth were then used to fnd the total selecton CLrv'es for all methods of attacbment, followng the procedure descrbed n the Appendx. As/frst step a free hand curve Was' ftted to the ratos betvveen the numbers of fsh meshed and those attached n other ways. Ths curv'e has a parabolc shfl.po wth a mnmum approxmately at the length of greatest selecton by meshng (Fgure 2). Ths show< that other methods of attachllent are of sgnfcance manly for the small and very large fsh and thus tend to ncrease the effcent selecton range of the net,.e. make the selecton curve more flat-topped. To fnd the selecton curv'e for the 6 hemp net the length dstrbuton of the meshed fsh taken by ths type of net was'compared by that of the 6 nylon net as adjusted for the effect of mesh selecton. Thsreafter the total selseton curve was found n the same way as for the nylon nets. Dscusson From Fgure t appears that the plots of the log. ral:;los devate consderably from lnearty, as would be expected because of the heterogenety wth regard to attacbment method and the relatvely small number of observatons consderng the very great szo range. Nevertheless, there s no great dfference bet'tjeen the three ndepenjent values of K, and they compare farly well wth the rato of ho.lf' the grth to the total length, whch was found to equal 3.04. To further test the valdty of the assumpton that the Leo.n selecton length s proportonal to the mesh sze, the lengths at whch the log. ratos equal zero, were plotted aganst the sum of the mesh sze as descrbed by Olsen (959). The plots ft farly well to a straght lne through the orgn, wth a slope of L599. Ths corresponds to a value of K ~ 3.98, aganst ~ 3.204. - The total selecton curv'es for the four df'ferent types of nets used are shown n Fgure 3. The cui'lres are farly flat-topped, :l!.e. the halbut gll nets are effectve over a great range of fsh sze. TIms the nylon nets have an eff'cency of 50% or more, as referred to~that of the mean selecton length, over'a range of about 04 cm, 0 cm and 0 cm for the 6!!, 9!! and 23 Tl uesh sze respectvely. The selecton curve for the hemp net s more peaked and the' 50% selecton range equb.ls about 83 cm, whch s 20% less tho,n -that of the 9! nylon net.
- 3 - It s a general experence n all gll-nej fsheres that nets made of polyamde fbres are more effcent than smlar nets madejratural fbres~ as for nstance stated by Sfftersdal (957). The results of these experments would suggest that at least part of ths-dfference n fshng power s caused by the extended selecton range of polyamde nets. Summary Mesh selecton data collected by expermental fshng wth halbut gll nets of dfferent mesh szes were analysed by the method descrbed by Holt (957). The rato between the mean selecton length for meshed fsh and the mesh sze Was found to be farly low, v~z. K = 3.204. A method for fndng an emprcal selecton curve descrbng both the selecton by meshng and all other knds of attachment was dev-eloped and appled to the data. The resultng total selecton curv'e for halbut gll nets was found to be relatv'ely flattopped, but less pronounced n nets made of hemp than of nylon. It s suggested that ths dfference n selecton range accounts for at least part of the dfference n effcency generally found between gll nets made of natural fbres and polyamde nets. References Halt, S. J. 957 Olsen, Stenar 959 Sfftersdal, G. 957 ta method of determnng gear-selectv'ty and ts applcaton!!. Paper No. S.5, Jo~nt ScL Meetng ICNAF /ICES/FAO, Lsbon, 957. (Ml:m.eographed) UMesh selecton n herrng gll nets!!. J.Fsh.Res. Bd. Canada, 6(3), 959.!lOn the fshng power of nylon gll netstl. Paper No. E.7, Jont ScL Meetng-ICN.A.F/ICES/FAO, Lsbon, 957 (Mmeographed).
Table 2. Halbut meshed by operculum and pont of greatest grth. Length dstrbuton and log.ratos when adjusted for dfferences n effort. -- ~--=7 no~;l~~~[~ =--=l-lj~~~-~~hodf ~~~!~s2;~sh:t---- -...-- -I -----I, r--... Range n om ~~-- ~ --. -nl-~~~f. f~:..] -<~~r ~~ - lloge B/A I logo C IB I logo CIA ~ L effort effort, I -------- --------- - -- - - ----- - ---- ---- ---------- ------------- ---- ------- -----f- --------------- ~--- L_ -------- -----j 70-89 79.5 I 5 4 5 84 I I I I 90-09 99.5 3 5 7.30 2 3.22-0.57' I I -.3956 0--29 9.5 44 6.06 30-49 39.5 34 24 35.04 I 6 9.66 -l.do79 I -0.5084 I I 5 8.05 0.0296 -.4708 I -. 562 -.4407 50-69 59.5 24 2 30.66 7.27 0.2450 -.0009-0.7559 70-89 79.5 I 4 6.06 8 28.98 0.37 0.5902 0.7275 90-209 99.5 I 5 8.68 9 30.59 0.8475 0.9628.82 20-229 29.5 I 4 7 0.22 I 2 9.32 0.938 0.6369 230-249 239.5 2,6 I I!= 2~ l j, 7.27. 5748 ~ ~~~~. J.
- 5- APPENDIX Method to fnd an emprcal total selecton curve descrbn~ all means of attachment. If the selecton cun'e for one method of attacbment (Le. meshng at the maxllae, operculum, pont of maxmum grth etc.) s known, an emp,rcal selecton curv'e descrbng all knds of attachment can' be found, although no smple mathematcal expresson for ths curve may be establshed. Let Yl be the ordnate of the known selecton cunre at length, nl the catch n number of fsh caught by ths method of attacbment. Smlarly Yl'': and nl l refer to the selecton cun'e and catch of all other methods of attachment. Then, P () and P Y t (2) where N_ s the total number of fsh of length,. encountered by the net fu~d P s a constant. We hal'e now n Y Y ~ (3) and Y t n! n_ J. (4) and the ordnate for the total selecton curve t. t == Yl + Y t == " Y l ( + n l! n_. )... 00... 000. (5) or, f we chose Y l as unty ==. m y Y r Im ( + ) 0.00 0... 0000 (6)
- 6 - loge ~~~ do~!/.8 ~ A )/.6.4.2.0 r la S/ L / j~:~::'/ :! / J'." '-- m.8 0.6 00. U ::) -l'- + 0.4 0.2 0.0 0.2 004 0.6 0.8 Lo.2.4.6 ~, _/."IJt,:~, I!! LJ f".! ""' j I -;;o----2o-----4;---:':~:,'"~---"":t-"--_;.7so_---'2oo----z2"6 ~///.«,.ff" L. """"""""""./~.,-,!~ " "~"._ 0 """""",,()"""""" f",,,/f! "'9.~""'" : """ -).~~"er! "'"""""". L ; I I - L~ r- I j~~!./.! Fgure. Plots of log. ratos ~ganst length log e log e () log e ( I,entt--\"cm) BI-f\. C/I'!. C/B
- 7 -.6... -... :... -... -... "... -.... "',)...} () ---' ''o"''''''''''''''''',,,,,,.,q,,,,,,,,,,,,,,. """,,,,,,,,,,,,,,,,,,,,,,,,,""t''.t''''''/'"'''''''' f!~'i"'ii'ii!0ht"'nll'ii\i'\t\\t\h\\"'''\\'\\ f, I --L-, ~I _---_ j ---!.--,_ 80 loo 20 40 60 80 200 220 240 Length (cm) Fgure 2. Ratos between numbers of fsh meshed and those attached n other ways, 6!! - nylon nets 0!, lod- 0.9t-.! 008 0.7~ 0.6I - N 23! j"",/ H 6 ""'" "'>,. :: f"",,,/'/'/ """"''''''''t'''''''''''''''''''''''''''''''''/''/''// """,.,''''''''''"."."."".".", "", """""""''''''''''''''''''''''.Jl: 9 o 0 ]-:",...""...,,""'"'''''''''' : I : t, t I ~ : ~ l L_L--. l...j.._,_---l.---.j L. I - L,_--'- ; 80 loo 20 40 60 80 200 --=2-':2-0- """"""',, ;f. 6 240 Length (cm) Fgure 3. Relatv'e selecton curv'es descrbng all methods of attacbment, H 6 - hemp 6!!, }IT 6 - nylon 6T!, N 9 - nylon Hll, N 23 - nylon 23!.