SWORDFISH DEAD DISCARDS AND LIVE RELEASES BY JAPANESE LONGLINERS IN THE NORTH ATLANTIC OCEAN IN

Transcription

1 SCRS/2003/134 Col. Vol. Sci. Pap. ICCAT, 56(3): (2004) SWORDFISH DEAD DISCARDS AND LIVE RELEASES BY JAPANESE LONGLINERS IN THE NORTH ATLANTIC OCEAN IN Kotaro Yokawa 1, Takumi Fukuda 2 SUMMARY The Japan fishery management authority instructed Japanese longliners operating in the northern stock area (north 5ºN in the Atlantic) to discard all their swordfish catch from the beginning February 2000 to avoid over quota. In the present study, the number and weight swordfish discarded and d by Japanese longliners in the north Atlantic for the period were estimated using data from the discarded and d number obtained every 10 days and by vessels, and size information obtained by observer records, data from longline research cruises and reports from commercial vessels. RÉSUMÉ Les autorités japonaises responsables de la gestion des pêcheries ont donné pour instructions aux palangriers japonais qui opèrent dans la zone du stock nord (nord de 5ºN dans l Atlantique) de rejeter toutes leurs captures d espadon à partir du mois de février 2000 dans le but d éviter un dépassement de quota. La présente étude fournit une estimation du nombre et du poids de l espadon rejeté et remis à l eau par les palangriers japonais dans l Atlantique nord pour la période , en utilisant le nombre de rejets et de remises à l eau obtenu tous les 10 jours et par bateau, ainsi que les données de taille consignées dans les carnets des observateurs, les données recueillies par les campagnes de recherche palangrières et les registres des navires commerciaux. RESUMEN La autoridad de ordenación pesquera de Japón instruyó a los palangreros japoneses que operan en la zona del stock septentrional (al norte de 5º N en el Atlántico) para que descartasen toda su captura de pez espada, desde comienzos de febrero de 2000, para evitar superar la cuota. En este estudio se ha estimado el número y peso del pez espada descartado y liberado por los palangreros japoneses en el Atlántico norte durante el periodo , utilizando los datos de descartes y liberaciones en número, obtenidos cada 10 días y por barco, y la información de talla obtenida de los registros de los observadores, los datos de los cruceros de investigación de palangre y los informes de los barcos comerciales. KEY WORDS Swordfish, Long lining, Fish catch statistics 1 National Research Institute Far Seas Fisheries, 5-7-1, Shimizu-orido, Shizuoka , Japan, yokawa@affrc.go.jp 2 Japan Fishery Agency,1-2-1 Kasumigaseki, Chiyoda-Ku, Japan

2 1 Introduction Japanese longliners started to discard all their catch dead swordfish and living ones (Yokawa and Fukuda, 2001). At same time, fishermen stopped to report their swordfish catch in the north Atlantic by logbook system as it was only used to record their landing. Japan fishery management authority has collected statistic and biological information discarded/d swordfishes by several ways such as the radio reporting system and observer programs. This study reviews these data for discarded/d swordfishes from the view points for their use in the study for stock assessment. 2 Materials and methods 2.1 Estimation number dead discarded and live d fish When commercial longliners started to discard/ all their swordfish catch in the north Atlantic, Japan fishery management authority demanded them to report number their discarded and d swordfish by every 10 days period by a radio reporting system. These data are used in the estimation number yearly total discarded and d swordfish. Though Japan fishery management authority asked all commercial longliners operated in the north Atlantic to submit numbers their swordfish discards and s, part them have not their records. Yearly total amount discard/d swordfish by Japanese longliners are estimated by the raising method described below. The data is aggregated by year, quarter, and area to raise to the total using the same area stratification as estimation sex-ratio key swordfish (Fig. 1, Mejuto et al., 1997). In relatively small cases where commercial vessel operated in more than one area in a 10 days period, the data the vessel for that period is not used for the estimation. Aggregated numbers discarded and d swordfishes are extrapolated using ratio between sum commercial vessels which report discard/d records and total amount which calculated using log book data (Tables 1 3). Stratum which has no discard and in a 10 days report but corresponding log book information shows there are operations, numbers discarded and d swordfish are extrapolated using information nearest strata, assuming they probably actually caught some swordfishes (Table 4). Total number discard and a stratum whose number discard and is 0 but these report with small amount (< 20,000 hooks) is also estimated in a same way as described above, as discard and data with small amount is doubtful to be used as a raising base. 2.2 Estimation total weight dead discarded and live d fish Total weight dead discarded and live d fish is estimated by the catch at size discarded/d swordfish by year, quarter, and area. The catch at size is estimated by the same method in Yokawa et al. (2002). Source size information are observer records, data from longline research cruises and reports from commercial vessels. Many commercial vessels measured and size data dead discarded swordfish. They are mostly whole weight, but some significant numbers fish are measured in length, therefore all weight data are converted into lower jaw fork length (LJFL) using conversion factors in Anon. (1988). Incidentally, such weight data has been kept some fishermen for their own information before discard. Size data (LJFL) are aggregated in the same way as the discarded and d number swordfish, and used for the estimation catch at size by year, quarter, and area. For strata with not enough number size data are available (Table 8), it was substituted by the length frequency data nearest strata, which is chose by following priority; a) Nearest quarter in the same area and same year. b) Sum length frequency 4 quarters in the same area and same year. c) Nearest year in the same area and same quarter. d) Nearest year in the same area but different quarter. e) Sum length frequency multiple years in the same area and same quarter. f) Sum length frequency 4 quarters multiple years in the same area. Length frequency by year, quarter, and area are raised by the estimated number discarded and d 968

3 swordfish in each stratum to obtain catch at size. Total catch weight, total discarded weight, and total d weight in each stratum is calculated by the estimated catch at size and length-weight relationship showed in Anon. (1988). 3 Results and Discussions 3.1 Estimation number dead discarded and live d fish Tables 1 3 shows number discards and s by year, quarter, and area, and Table 5 shows summary these data. Values in 1st quarter in 2000 do not contain those in January, as discarding/releasing started in February The data discards and s sent by commercial vessels covered % total, and % total number the stratum (Table 5). Lower coverage discard/ report in 2002 than previous two years would mainly caused by the introduction a new reporting system enacted by Japan fishery management authority, which demand the vessels to report their discard/ data using same type sheet as log-book. The new reporting system is necessary to collect size information dead discarded fish used in the stock assessment. Though the reporting system started in August 2002, and it has been used jointly with the former 10 days period by the radio reporting system, reporting ratio the new reporting system is not high so far (approximately less than 50%). It seems that this new reporting system has not well filtered into the fishermen. Some complicating situation about reporting swordfish discards/ may be occurring in fishermen. Extrapolations in estimating number discarded/d swordfish by stratum were made in 3 strata in 2000, 7 strata in 2001, and 11 strata in 2002 (Table 4). Lower coverage size data in 2001 and 2002 could be reduced in near future by using size data by the new log-book system, which not compiled yet. Quality size data in the new log-book system is apparently better than those used in this study, because fishermen tend to ignore small sized fishes (Yokawa, 1999), and more than half size data used in this study by fishermen. Ratios live d fish to the total were 30% - 40% and no large differences or consistent tendencies were seen among different years, quarters, and areas (Tables 1 3, 5, and 6). 3.2 Estimation total weight dead discarded and live d fish Based on the estimation discarded/d number swordfish by year, quarter, and area, and estimated length frequencies, weight discarded/d swordfish were estimated (Table 6). Coverage size data in the estimation catch at size were not so high (Table 8), and about number size data for about half strata were not enough in estimation and substituted by other data. Such situation is common in Japanese data and coverage size data in but is still better than those before the 1994 when size data coverage was lower than 8% (Yokawa, 1999). Estimated total discard/d weights in 2000 were bit larger than those in last year (778 ton, submitted as TASK-I data). This mainly comes from differences methods estimation between this study and last year. In last year, the discard/d weights calculated as the product yearly total catch number and average weight, which was obtained by past log book data. Average weight used in last year was roughly 10kg light (70kg) than those in this year. In this study, average weight estimated through estimation catch at size. These differences size data come from increase average size in recent years (average weight used in last year was those before February, 2000 when strong year classes were young). 4 Conclusion Number and weight swordfish discarded and d by Japanese longliners in the north Atlantic for periods were estimated by year, quarter, and area in this study. Discarded and d number could be estimated using data from 10 days period by the radio reporting system which has higher coverage. 969

4 Though the coverage size data used in estimation discarded and d weight is lower than the one recommended by ICCAT SCRS Swordfish Working Group (minimum 20%, Anon. 1999), they are still higher than those before Therefore it is concluded that dead discarded and live weight estimated in this study are best available data and they can be used for the input TASK-I as well as input for stock assessment. References ANON Report the ICCAT swordfish workshop. ICCAT, CVSP. Vol. XXVII, ANON Report the ICCAT data preparatory meeting for south Atlantic swordfish. ICCAT SCRS DOC SCRS/99/19, 71p. MEJUTO, J., J. M. De La Serna, and B. Garcia Some considerations on the spatial and temporal variability in the sex ratio at size the swordfish, Xiphias gladius. ICCAT SCRS DOC SCRS/97/32. 12p. YOKAWA, K. and Fukuda Recent Status the Swordfish Catch by the Japanese Longliners in the Atlantic Ocean. ICCAT SCRS DOC. SCRS/2001/147, 3 p. YOKAWA, K., Y. Takeuchi and K. Uosaki Standardization age specific CPUE swordfish caught by Japanese longliners in the north Atlantic. ICCAT SCRS DOC, SCRS/2002/144. 5p. YOKAWA, K Recent status the catch swordfish by the Japanese longliners in the Atlantic Ocean. ICCAT SCRS DOC SCRS/99/126, 9p.. 970

5 Table 1. Discard and report by quarter and area in Values in 1st quarter in 2000 do not contain those in January, as discarding/releasing started in February year quarter area total am ount am ount discard ratio estim ated discard estim ated coverage CPUE (n/1000 hooks) ,794 30, % ,704 91, % % ,891,915 2,454, % % , , % % ,472,057 1,532, % % ,139 17, % % ,338 53, % % ,211, , % % , , % % , , % % , , % % , ne ne 0% ne ,569,764 4,068, % % , , % % ,215, , % % ,217 18, % % , ne ne 0% ne ,343, , % % , , % % , , % % ,467 51, % ,447,662 3,663, % % ,008 73, % ,951,710 4,194, % % ,710 78, % % ,378 82, % ,286 20, % % ,154, , % % , , % % , , % % ,095 49, % % ,309,599 1,121, % % ,752,030 4,683, % % , , % % , , % % ,250,621 1,827, % % ,108 48, % % ,237 48, % % ,109 19, % ,735,840 3,666, % % , , % % , , % % , , % %

6 Table 2. Discard and report by quarter and area in year quarter area total am ount am ount discard ratio estim ated estimated discard coverage CPUE (n/1000 hooks) , , % % , , % % ,436 67, % % ,971,063 4,001, % % ,177, , % % ,197,025 1,595, % % ,018 5, % , , % % ,269,688 2,827, % % , ne ne 0% ne , , % % , ne ne 0% ne , , % % , ne ne 0% ne ,531,901 2,504, % % , , % % , , % % , ne ne 0% ne , , % % ,876,422 1,577, % % , , % % ,405 88, % % ,065 62, % % ,999,192 3,462, % % ,651 30, % % ,805,068 2,447, % % ,082 45, % ,421 11, % % , , % % , , % % ,931 85, % % ,524 57, % ,196, , % % ,497,388 2,636, % % , , % % ,284 16, % % ,620,005 1,975, % % , ne ne 0% ne ,067 61, % % ,643 14, % ,885,749 5,584, % % ,236, , % % , , % %

7 Table 3. Discard and report by quarter and area in year quarter area total am ount am ount discard ratio estim ated discard estimated coverage CPUE (n/1000 hooks) ,772 35, % % , , % % ,825 87, % ,555,829 3,452, % % , , % % ,003, , % % ,262 5, % , , % % ,228,113 3,561, % % , , % % , ne ne 0% ne , , % % , ne ne 0% ne ,737,058 1,808, % % , , % % , , % % , ne ne 0% ne , , % % ,327, , % % ,492 8, % ,226,477 1,496, % % ,017,766 1,952, % % ,982 10, % % ,225 52, % , ne ne 0% ne ,048 12, % , ne ne 0% ne , ne ne 0% ne ,537 28, % ,075,830 2,409, % % ,924 75, % ,513 74, % ,183, , % % ,000 18, % % , ne ne 0% ne ,703 2, % % ,301, , % % ,386, , % % , , % ,841 17, %

8 Table 4. Summary substitution in estimating total number catch (discarded and d) swordfish strata with no or few observations. year quarter area total hook hook Stratum used for substitution year quarter area CPUE (n/1000 hooks) estimated discard and num ber estimated num ber

9 Table 5. Summary tables 1 3 (containing extrapolated values). Values in 2000 do not contain those in January as discarding started in February. year coverage coverage ratio average C P U E 1) stratum 1) catch 2) (n/1000 hooks) % 93% % % 93% % % 73% % ) Data stratum with few observations are not included. 2) Calculated by the quotient between total number catch and. 3) Average value CPUE each stratum. Data stratum with few observations is extrapolated. Table 6. Estimated numbers and weights the sum discard and by area and year. For the estimation weight d fish, same length frequency as discarded fish is assumed. Values in 2000 do not contain those in January as discarding started in February. Discard and w eight (ton) T otald w eight Ratio relesed fish Area Area Area % 29% 25% % 35% 48% % 40% 39% % 46% 41% % 0% % 37% 45% % 37% 37% % 32% 39% % 38% 37% % 39% 34% % 37% 39% % 28% 50% % 34% 35% total total total 36% 37% 41% Discard and num ber Total num ber Ratio relesed fish Area Area Area % 29% 17% % 35% 48% % 40% 40% % 46% 41% % 33% % 37% 45% % 37% 37% % 32% 39% % 38% 37% % 39% 34% % 37% 39% % 28% 50% % 34% 35% total total total 35% 37% 41% 975

10 Table 7. Number size data, and number sum discards and s for by year and by area.. size data discard and area area total total Table 8. Coverage size data. Values showed in bold type were used for estimation catch at size, and values showed in normal type indicated that size data was substituted. area % 7% 67% 2 5% 7% 1% 3 23% 7% 15% 4 8% 20% 27% 5 40% 67% % 7% 5% 7 27% 8% 8% 8 20% 15% 38% 9 32% 10% 5% % 22% 4% 12 4% 5% 4% 13 2% 1% 0% 14 12% 5% 4% total 12% 9% 7% 976

11 Figure 1. Area designation used in this study (sited from Mejuto et al., 1997). In this study, area 91, 92, 93, and 17 were converted into area 11, 12, 13, and