SCRS/2011/198 Collect. Vol. Sci. Pap. ICCAT, 68(5): (2012)

Transcription

1 SCRS/2011/198 Collect. Vol. Sci. Pap. ICCAT, 68(5): (2012) PRELIMINARY VIEW OF BY-CATCH HOTSPOT: DISTRIBUTION OF SEABIRDS FROM TRACKING DATA, INTERACTION MAP BETWEEN SEABIRD DISTRIBUTION AND LONGLINE EFFORT AND BY-CATCH DISTRIBUTION IN THE ICCAT CONVENTION AREA OF THE SOUTHERN HEMISPHERE Yukiko Inoue, Kotaro Yokawa, Hiroshi Minami, Daisuke Ochi 1 SUMMARY Seabird distribution maps are presented based on seabird tracking data, interaction maps between longline fishery effort and seabird distribution, and data on the distribution of by-catch CPUE of seabird species in the South Atlantic, based on Japanese by-catch data, to identify by-catch hotspots. Tracking data indicate the highest concentrations of seabird breeding distribution in the area between 5-10W, 35-40S, and 35-40S, 10W-15E, and also 35-60S, 55-65W during non-breeding. The available by-catch data confirm the distributions shown by the tracking data, but there are exceptions in each species and the degree of concentration of seabird distribution did not necessarily agree with the degree of CPUE for each species. The degree of interaction data was low level for the latitude 40-45S while CPUE of by-catch data was quite high. Interaction data showed concentration for latitudes 25-40S longitudes 55-40W, where there were no by-catch data. It was suggested that the three methods should be integrated to define the hotspot. Distribution of by-catch CPUE in albatrosses was high especially off South African waters and in the southeastern Indian Ocean. Thus, these two areas and the SW Atlantic would be considered as risk areas for seabird by-catch, and it is necessary to introduce appropriate mitigation measures there. RÉSUMÉ Ce document présente des cartes de distribution des oiseaux de mer basées sur les données de suivi des oiseaux de mer, les cartes d'interaction entre l'effort de pêche palangrier et la distribution des oiseaux de mer, et les données sur la distribution de la CPUE des captures accessoires d'espèces d'oiseaux de mer dans l'atlantique Sud, sur la base des données de prises accessoires du Japon, en vue d'identifier les points névralgiques des prises accessoires. Les données de suivi indiquent les plus grandes concentrations de la distribution des oiseaux de mer reproducteurs dans la zone comprise entre 5-10W, 35-40S, et 35-40S, 10W-15E, et également la zone 35-60S, 55-65W en dehors de la saison de reproduction. Les données de prises accessoires disponibles confirment les distributions montrées par les données de suivi, mais il existe des exceptions pour chaque espèce et le niveau de concentration de la distribution des oiseaux de mer ne coïncidait pas nécessairement avec le niveau de la CPUE pour chaque espèce. Le degré d'interaction était faible pour la latitude 40-45S, tandis qu'il était assez élevé pour les données de CPUE des prises accessoires. Les données d'interaction ont fait apparaître des concentrations pour les latitudes 25-40S et longitudes 55-40W, où il n'y a pas de données de prises accessoires. Il a été suggéré que les trois méthodes devraient être intégrées afin de définir le point névralgique. La distribution de la CPUE des captures accessoires d'albatros était élevée, notamment au large des côtes sud-africaines et dans le Sud-Est de l'océan Indien. C'est pourquoi ces deux zones et l'atlantique Sud-Ouest seraient considérés comme une zone à risques pour les prises accessoires d'oiseaux de mer, et il est nécessaire d'y introduire des mesures d'atténuation appropriées. RESUMEN Se presentan los mapas de distribución de aves marinas que se basan en los datos de seguimiento de aves marinas, en los mapas de interacción entre el esfuerzo de la pesquería de palangre y la 1 National Research Institute of Far Seas Fisheries Tuna and Skipjack Resources Division. 1784

2 distribución de aves marinas, así como en los datos de distribución de la CPUE de captura fortuita de especies de aves marinas en el Atlántico sur, basados en los datos de captura fortuita japoneses, para identificar los puntos álgidos de capturas fortuitas. Los datos de seguimiento indican que las concentraciones más elevadas de distribución de reproducción de aves marinas se sitúan en la zona entre 5º-10ºW, 35º-40ºS y 35º-40S, 10ºW-15ºE, y también entre 35-60S, 55-65W durante la temporada en que no se reproducen. Los datos de captura fortuita disponibles confirman la distribución obtenida de los datos de seguimiento, pero se han observado excepciones en cada especie y el grado de concentración de la distribución de aves marinas no coincidió necesariamente con el grado de CPUE para cada especie. El nivel de datos de interacción fue bajo para la latitud 40º-45ºS, mientras que los datos de CPUE de captura fortuita fueron bastante elevados. Los datos de interacción mostraban concentraciones en las latitudes 25º-40ºS y longitudes 55º-40ºW, en las que ho había datos de captura fortuita. Se sugirió que los tres métodos deberían integrarse para definir los puntos álgidos. La distribución de la CPUE de captura fortuita en albatros fue elevada sobre todo en aguas frente a Sudáfrica y en el océano Indico suroriental. Por tanto, estas dos zonas y el Atlántico suroccidental deberían considerarse zonas de alto riesgo para la captura fortuita de aves marinas, y es necesario introducir las medidas de mitigación apropiadas en dichas zonas. KEYWORDS Seabird distribution, CPUE, by-catch 1. Introduction The problem of seabird by-catch in longline fisheries is thought one of the risks to seabird conservation (Brothers 1994). The necessity to conserve seabirds has led to the development of several kinds of by-catch mitigation technologies and these have been deployed by many pelagic longliners (Kiyota 2002, Yokota & Kiyota 2008). To avoid by-catch effectively, it is crucial to understand the areas where many seabird by-catches have occurred (= by-catch hotspot). There could be multiple ways to define the hotspots of seabirds by-catch and the well defined hotspots could be used as beneficial tools from the view points of scientific studies and conservation of seabirds. At the ICCAT meeting held last May, it was revealed that Japanese by-catch data indicate that the rate of seabird by-catch in the North Atlantic is lower than south (Inoue et al. 2011). And also it was discussed that the risk analysis of seabird by-catch should be integrated across the southern hemisphere as the distributing areas of seabirds by-caught in the South Atlantic extend over the other two Oceans. Thus, the by-catch data of Indian Ocean and Pacific Ocean were added in this study. In our previous document, it was found that the distribution pattern of CPUE of seabirds caught by Japanese longliners differ from the pattern of distribution probabilities of seabirds estimated by tracking data. The observed differences could be partially due to the lack area of observer data but could also be attributed to the behaviors of seabirds as distribution probabilities may not have strong linear relationship with the by-catch probabilities. In the present study, distribution pattern of seabirds by-catch and tracking data of major bycaught seabird species were compared to sketch the areas of hotspots of seabird by-catch for the tuna longline fisheries in the southern hemisphere. Seabird community, abundance and aggressiveness toward pelagic longline operations is geographically varied. In South African waters, it is reported that large numbers of seabirds aggregate around operating pelagic longline vessels and intensely attack baited hooks, attacks being made mainly by diving seabirds (Melvin et al. 2009). In the case of the North Pacific, the situation is very different in that the number of species aggregating and their attacking rate during the pelagic longline operation is quite a lot lower than those in South African waters, and there are no deep divers (Sato et al. 2010). Thus, seabird aggressiveness and by-catch risk differs between regions. The higher quality of the results of risk analysis, especially for the designation of hotspot should offer valuable information for the introduction of effective mitigation measures with lesser burden on fisherman. By-catch may occur in the area where seabird distribution overlaps with areas of high fishing effort. On the other hand, the probability of by-catch for each species will depend on its behavior and interaction with other seabirds, especially for the seabird species which have lower ability of diving to pick up underwater longline baits. The frequency of seabird by-catches would also be affected by the intrinsic demands of seabirds for food as well as the food availabilities. Therefore, for the designation of by-catch hotspots, it is essential to consider the seabird distribution from tracking data, interaction map which is overlap areas between seabird distribution and fisheries effort, and they need to be compared with distribution of seabird by-catch. 1785

3 At-sea distribution or foraging effort of seabirds changes seasonally according to phenology; laying an egg, rearing a chick and migrating. For example, during the chick-rearing period, a parent s foraging range is restricted in order to return to its colony for chick provisioning and foraging effort is increased in order to feed the chick. In the non-breeding period, some albatrosses and petrels migrate long distances. Also, the fishing ground of tuna longline fishery changes by season due to the seasonal migration pattern of target fishes or seasonal changes of target species. Thus, the actual position of by-catch hotspots would change seasonally. It is important to closely examine seasonal distribution of each seabird species. Our study focuses on seven seabird species in the South Atlantic, based on species of concern due to population decreases, or for which CPUEs were high in the Japanese by-catch data. Examination is made of 1) distribution from the tracking data, 2) interaction maps calculated with fisheries effort and seabird distribution 3) by-catch distribution using data collected by Japanese observers, fishing boat for high school training and chartered research boats in s010 in the southern hemisphere. The results of (1-3) are compared and the hotspots to introduce by-catch mitigation measures are discussed to be referred at the development of seabird conservation plans. 2. Methods 2.1 Tracking data Materials Tracking data used in this analysis were provided with permission of the data owners for the relevant data sets held within Global Procellariiform Tracking Database ( This builds on the previous analysis of tracking data for ICCAT (ACAP 2010), incorporating new data. The sample sizes are shown in Appendix 1, and data owners are indicated below each map. Data processing Methods follow those used in Alderman et al. (2011). We divide the data into the distribution of albatrosses and petrels during the breeding season and distribution during the non-breeding season. 2.2 Interaction index analysis Data processing The longline fishing effort is the average annual number of ICCAT longline hooks from per 5x5 grid square, based on the modeled effort database that the ICCAT Secretariat produced for the ICCAT seabird assessment to fill known gaps in catch and effort data. The % of seabird distribution per 5x5 grid square was multiplied by the average longline fishing effort per 5x5 degree square to give a measure of interaction (% species distribution x hooks, Tuck et al. 2011). 2.3 By-catch data Materials We used seabird by-catch data collected by science observers for commercial tuna longliners, training vessels of high schools, as well as chartered research boats. The majority of the data was collected by the first two data sources on a volunteer basis. The data from observer programs in the southern bluefin tuna operation were obtained during and the data from high school training vessels and chartered research boats were obtained during In the majority of cases, the quality of data on seabirds were maintained by giving special lectures to each observer, and species identification was conducted using photos taken by observers in the designated way, and descriptions in the species identification guide (Nakano 2002). In this study, the older classification system for the species identification of albatrosses was used because of the difficulty to identify some newly described albatross species in the most recent year. There are some difficulties existing for the species identification of seabirds based on the appearance (external morphology), especially for some albatross species. Though species identification of seabirds in the Japanese observer data had conducted by the scientists in National Research Institute of Far Seas Fisheries (NRIFSF) using photos taken by observer, NRIFSF is now plan to re-check species identifications in the Japanese observer data with the specialists of seabird taxonomy 1786

4 Data processing Using these data, the number of seabirds caught and number of hooks were summarised into a grid of 5 x5 respectively. The number of seabirds caught was divided by number of hooks and multiplied by 1000 to obtain average CPUE values of seabirds for each 5 x5 grid cell. The use of seabird by-catch mitigation measures changed by year, area, season and even by vessel and this is likely to have some large influence on the degree of seabird by-catch, but these effects are not taken into account in the calculation of average CPUE for each 5 x5 grid. For the species with higher occurrence in the South Atlantic, and vulnerable species, quarterly CPUE distributions (January-March, April-June, July-September and October-November) are presented. 3. Results 3.1 Seabird distribution from tracking data The distribution of all 7 species (black-browed albatrosses sooty albatrosses, Tristan albatrosses, wandering albatrosses, Atlantic yellow-nosed albatrosses, grey-headed albatrosses and white-chinned petrels) during breeding was in the central south Atlantic Ocean, while during non-breeding distribution was highest in sublittoral zones such as those off Argentina and off southern African waters (Figure 1, a,b). Distribution of black-browed albatrosses during breeding concentrated in Argentina waters and around the Drake Passage, whereas non-breeding birds used southern African waters as well as Argentinean waters and the Drake Passage (Figure 1 c,d). The main distribution of sooty albatrosses (both breeding and non-breeding) was in the pelagic waters of the central south Atlantic Ocean (Figure 2, a,b). The center of distribution of breeding Tristan albatrosses was in the central south Atlantic Ocean but non-breedingbirds ranged across the south Atlantic (Figure 2 c,d). Wandering albatrosses distributed off Argentina and off south African waters during breeding, while non-breeding birds ranged broadly in south Atlantic Ocean (Figure 3 a,b). Yellow-nosed albatrosses were distributed further north compared with other albatrosses during both breeding and non-breeding, focused in the south eastern Atlantic Ocean (Figure 3 c d). The distribution of breeding grey-headed albatrosses was further south than other species, and focused in the Drake Passage, around South Georgia and off South Africa, while non-breeding birds ranged broadly in the southwestern Atlantic Ocean (Figure 4 a b). Breeding white-chinned petrels distributed around South Georgia, while non-breeding birds distributed near the Argentinean coast (Figure 4 c,d)). Distributions of all these species during non-breeding were larger than ones during breeding (Figures 1-4). 3.2 Interaction maps from longline effort and seabird distribution The interaction map based on the combined breeding distribution of all 7 species shows highest overlap off Argentina water, 35-40S, 50-55W, and southeastern Atlantic Ocean, 30-40S, 10W-15E, while the combined interaction for non-breeding birds shows highest overlap in Argentina waters, 25-40S, 40-55W and the southeastern Atlantic Ocean, 20-45S, 10W-20E (Figure 5). The interaction of black-browed albatrosses during breeding was high off Argentina and during non-breeding was high off southern African waters as well as off Argentina water (Figure 6). The interaction of sooty albatrosses both during breeding and non-breeding was broadly high in southeastern Atlantic Ocean (Figure 7). The interaction of Tristan albatrosses during breeding was high in the southeastern Atlantic Ocean but also included some high interaction in the Southwest Atlantic, and the interaction map for non-breeding birds includes broader areas, particularly including the coast of South Africa and Namibia (Figure 8). The interaction of wandering albatrosses during breeding was high off Argentina, Uruguay and Brazil while during non-breeding, interaction was also high in southern African waters (Figure 9). High interaction of Atlantic yellow-nosed albatrosses both during breeding and non-breeding was highlighted from Angola, through Namibian to South African waters (Figure 10). Interaction of grey-headed albatrosses was concentrated below 35 degrees South, reflecting their more southerly distribution (Figure 11). Interaction of white-chinned petrels during breeding was focused around South Georgia and off southern African waters, while during non-breeding it had high interaction off Argentina, Uruguay and Brazil (Figure 12). 3.3 By-catch distribution Procellariiform seabirds caught by the longline sets monitored by the observers in the ICCAT area between included 2 genus, 9 species and 309 albatross individuals; 1 genus, 2 species and 34 individuals of giant petrels; and 3 genus, 5 species and 106 individuals of petrels. The total number of by-caught seabirds was 502 individuals (Table 1). Grey-headed albatross was the species most frequently caught in the ICCAT convention area in (n=90, Table 1), followed by black-browed albatrosses (n=63, Table 1) and white-chinned petrels (n=47, Table 1). Few Atlantic yellow-nosed albatrosses and sooty albatrosses were caught in the ICCAT Convention area; 3 individuals and 4 individuals, respectively (Table 1). 1787

5 The CPUE for all seabird species combined was highest off southern African waters in April to September, in southeastern Indian Ocean in September to December and off Chile in July to December (Figure 13). CPUE of albatrosses was high especially off African waters and in the southeastern Indian Ocean (Figure 14). The CPUE for giant petrels was lower than that of albatrosses or petrels and occurred in the southeastern Indian Ocean and off African waters but not off Chile (Figure 15). By-catch of petrels in the ICCAT convention area was higher than that in the WCPFC area and it appeared off African waters, off Chile and in the southeastern Indian Ocean (Figure 16). By-catch CPUE for black-browed albatrosses was high off South Africa in April to September and in the southeastern Indian Ocean in September to December (Figure 17). There were few records of sooty albatross: low CPUE occurred off South Africa and in the southeastern Indian Ocean (Figure 18). Wandering albatrosses were caught in a number of areas across the Southern Hemisphere, off South Africa, in the southeastern Indian Ocean, the Tasman Sea and off Chile (Figure 19). By-catch CPUE for yellow-nosed albatrosses was regionally high off southern African waters and in the south eastern Indian Ocean (Figure 20). By-catch of grey-headed albatross occurred below 35 degrees south, off southern African waters, in the southeastern Indian Ocean. None were reported caught in the South East Pacific (Figure 21). By-catch of grey petrels appeared mainly off Africa and Chile (Figure 22). By-catch CPUE of white-chinned petrels was high off South Africa and off Chile, and also occurred in the southeastern Indian Ocean. 4. Discussion 4.1 Comparison between seabirds distribution and by-catch distribution Based on available tracking data, the distribution of breeding adults for all seven species during the breeding season concentrated around their colonies, but the distribution of non-breeding birds included high concentrations in coastal areas, and the range of distribution was broadened overall. Tracking data indicate highest concentrations of albatross and petrel distribution between 5-10W, 35-40S during breeding, 35-40S, 10W-15E, and also35-60s, 55-65W during non-breeding. The distribution patterns estimated by the tracking data, of albatrosses and petrels analyzed in this study were varied by species. Black-browed and Atlantic yellow-nosed albatrosses and white-chinned petrels had particularly high concentrations near the coasts, especially for non-breeding birds whereas sooty albatrosses and Tristan albatrosses had highest concentrations in pelagic areas, and wandering albatrosses and grey-headed albatrosses ranged broadly (Figures 1-4). The available by-catch data confirm the distributions shown by the tracking data but there are exceptions in each species. The distribution probability of each species shown in Figures 1-4 was not always agree with the distribution pattern of its CPUE. In addition, the areas of highest concentration of seabird distribution did not necessarily agree with the areas of highest CPUE in each species. There are several possible causes of the disagreement: 1) Vulnerability against longliners. The behavior of some species makes them less vulnerable to longliners. Seasonal migration patterns of seabirds inferred by the tracking data of some species indicates that they are not so vulnerable to the longline gear as others. 2) Interaction with other species. As explained in the concept of a white-chinned petrels dominant system (Melvin et al. 2009), the co-existence of diving seabirds might promote by-catch. This should be the case especially for the species with lower abilities of diving such as albatrosses. 3) Season. The tracking data presented here are divided into breeding and non-breeding season. However, it would be useful to compare the seasonal distribution of birds with the seasonal CPUEs (ACAP, 2010). 4) Difference in age and experience. The season (breeding or non-breeding) of by-catch data is estimation from the month when the individual was by-caught. The by-catch number of grey-headed albatrosses, shy albatrosses and white-chinned petrels in May, which is post-fledgling period, was the highest among seasons so many fledgling might be caught. 5) Lack of tracking data. Some gaps remain in the tracking data and the distribution of individuals in the other colonies might not be reflected in the seabird distributions shown here. For example, the tracking data sample size for sooty albatrosses from Indian Ocean colonies remains low, and the range of distribution is likely to be under-estimated. 6) Gaps in by-catch data. The available by-catch data do not cover all areas and seasons. 1788

6 4.2 Differences between the interaction maps and by-catch distribution The interaction maps for the 7 species combined predict high interaction areas (in terms of potential for highest number of birds killed) below 25S off Brazil, and below 15-20S off Namibia and South Africa. There are no by-catch data available from the South West Atlantic, and there are few by-catch data in the South East Atlantic for the latitudes 15-30S, therefore a comparison is difficult. However, there are available data for the latitudes 30-45S for the comparison, and the degree of interaction data was low level for the latitude 40-45S while CPUE of by-catch data was quite high. This would be attributed the difference between the seabird distribution and the by-catch distribution. In addition, because the interaction maps are predicting highest numbers of birds killed (which could include areas of low CPUE but high effort), whereas the by-catch data is more focused on high CPUEs. Thus, consideration is necessary when one interprets interaction data. However, interaction data can estimate potential areas of high interaction where there is no by-catch data. The previous document (Inoue et al. 2011) had a review that the by-catch level at south western Atlantic, off South America, could not evaluated due to lack of observer data. That area includes key concentrations of albatrosses and petrels, including those considered particularly vulnerable, and colonies of species including wandering albatrosses, black-browed albatrosses, grey-headed albatrosses and white-chinned petrels exists there. Thus, it is important to consider the overlap between seabird distribution and longline effort (as shown in the interaction maps) in these areas, and to consider sources of by-catch data from other fleets. 4.3 Preliminary view of by-catch hotspots The May 2011 meeting of the ICCAT Sub-Committee on Ecosystems discussed a proposal that an integrated approach should be taken to consider by-catch distribution patterns in the Southern Hemisphere, covering the Atlantic, Indian and Pacific Oceans. Data available from Japan identify by-catch of albatrosses in the southern hemisphere concentrated off southern African waters, especially in the SE Atlantic between April to September, and in the southeastern Indian Ocean in April to December. By-catch of petrels was concentrated in the south east Pacific. Based on the interaction maps, it was revealed that probability of overlap between fishery and seabird is high in the SW Atlantic. Consequently, off southern African waters, in southeastern Indian Ocean, and the SW Atlantic would be considered as risk area for seabird by-catch, and it is necessary to introduce appropriate mitigation measure there. 4.4 Distribution pattern of white-chinned petrels and mitigation measure According to tracking data and observer data, the distribution of white-chinned petrels was restricted to the SW and SE Atlantic in the ICCAT Convention area. Because white-chinned petrels bring baited hooks on the sea surface by diving to deep water (Mervin 2009), it is necessary to introduce mitigation measures to sink the bait to deep water before the petrels can catch the bait. The appropriate mitigation measure such weighted branch line should be introduced to the areas in which white-chinned petrels are distributed. 4.5 Future issues In our study, the occurrence of by-catch would be affected other factors as well as seabird distribution. To make use of tracking data to determine the by-catch hotspot, it is necessary to understand the factors affecting the occurrence of by-catch such as behavior of each species against longliners, composition of by-caught species, age of by-caught species, and differences between by-catch hotspot and ecological hotspot. Acknowledgements We thank for Cleo Small (Bird Life International) who manage collecting the permission from data holders and gave useful comments to draft of this document and Phil Talor (Bird Life International) who collect the sample size and name and affiliation of data holders. And we also thank for Richard Phillips (British Antarctic Survey), Dave Watts (Australian Antarctic Division), Flavio Quintana (Wildlife Conservation Society, Argentina), Graham Robertson (Australian Antarctic Division), Henri Weimerskirch (CEB CNRS, France), Javier Arata (Instituto Antártico Chileno), Jose Pedro Granadeiro (CESAM, Museu Nacional de História Natural, Portugal), Nic Huin (Falklands Conservation), Rachael Alderman (Australia), Rosemary Gales (DPIPWE, Australia), Samantha Petersen (WWF South Africa) who permit for us using valuable tracking data. 1789

7 References Alderman, R,D., Anderson, J., Arata, P., Catry, R., Cuthbert, L., Deppe, G., Elliot, R., Gales, J., Gonzales Solis, J.P., Granadeiro, M., Hester, N., Huin, D., Hyrenbach, K., Layton, D., Nicholls, K., Ozak, S., Peterson, R.A., Phillips, F., Quintana, G.R., Balogh, C., Robertson, G., Robertson, P., Sagar, F., Sato, S., Shaffer, C., Small, J., Stahl, R., Suryan, P., Taylor, D., Thompson, K., Walker, R., Wanless, S., Waugh, H., Weimerskirch, H , Albatross and giant-petrel distribution across the world s tuna and swordfish fisheries. Paper submitted to the Third Joint Meeting of the Tuna RFMOs. La Jolla, CA, July ACAP 2010, Albatross and petrel distribution in the Atlantic Ocean and overlap with ICCAT longline fishing effort. Paper submitted to the ICCAT Sub-Committee on Ecosystems, (SCRS/2010/050). Brooke, M. 2002, Albatrosses and Petrels across the World, Oxford University press, Oxford. Brothers, N. 1994, Fisheries should catch fish, not birds. Pandani Publisher, Hobart Inoue, Y., Yokawa, K., Minami, H., Ochi, D., Sato, N., Katsumata, N., 2012, Distribution of seabird by-catch using data collected by Japanese observers in in the ICCAT area. Collect. Vol. Sci. Pap. ICCAT, 68(5): Kiyota M. 2002, Incidental take of seabirds in longline fisheries: Nature of the issue and measures for mitigation, J Yamashina Inst Ornithol 34: Melvin E., Guy, T. and Read, L.B. 2009, Shrink and Defend: A comparison of two streamer line designs in the 2009 South Africa tuna fishery SBWG-3 Doc. Nakano, H. 2002, The Species Identification Guide for the Pelagic Longline fishery. National Research Institute of Far Seas Fisheries. Shizuoka. Tuck, G.N., Phillips, R.A., Small, C., Thomason, R.B., Klaer, N.L., Taylor, F., Wanless, R.M. and Arrizabalaga, H. 2011, An assessment of seabird - fishery interactions in the Atlantic Ocean. ICES Jour. Mar. Sci. 68: Sato, N., Ochi, D., Minami, H., Shono, H. and Yokawa, K (2010) Experimental comparison among four types tori-line designing in the Western North Pacific. WCPFC-SC /EB-WP-02 Yokota K. and Kiyota M. (2008) Mitigation measures to reduce incidental catch of seabirds: Recent developments in Japanese tuna longline fishery. Nippon Suisan Gakkaishi 74:

8 a) All 7 species during breeding b) All 7 species during non-breeding c) Black-browed albatrosses during breeding d) Black-browed albatrosses during non-breeding Figure 1. Distribution of seabirds in ICCAT area. Color legend shows the utilization distribution (%) calculated from tracking data. Circle plot shows the bycatch CPUE calculated at each 5x5 grid square from observer data in Black-browed albatross tracking data were provided for the purposes of this analysis with permission from Richard Phillips (British Antarctic Survey), Dave Watts (Australian Antarctic Division), Flavio Quintana (Wildlife Conservation Society, Argentina), Graham Robertson (Australian Antarctic Division), Henri Weimerskirch (CEB CNRS, France), Javier Arata (Instituto Antártico Chileno), Jose Pedro Granadeiro (CESAM, Museu Nacional de História Natural, Portugal), Nic Huin (Falklands Conservation), Rachael Alderman (Australia), Rosemary Gales (DPIPWE, Australia), Samantha Petersen (WWF South Africa). 1791

9 a) Sooty albatrosses during breeding b) Sooty albatrosses during non-breeding c) Tristan albatrosses during breeding d) Tristan albatrosses during non-breeding Figure 2. Distribution of seabirds in ICCAT area. Color legend shows the utilization distribution (%) calculated from tracking data. Circle plot shows the bycatch CPUE calculated at each 5x5 grid square from observer data in Sooty albatross tracking data were provided for the purposes of this analysis with permission from Henri Weimerskirch (CEB CNRS, France), Richard Cuthbert (Royal Society for the Protection of Birds, UK), Ross Wanless (BirdLife South Africa). Tristan albatross tracking data were likewise provided with permission from Richard Cuthbert (Royal Society for the Protection of Birds, UK), and Ross Wanless (BirdLife South Africa). 1792

10 a) Wandering albatrosses during breeding b) Wandering albatrosses during non-breeding c) Atlantic yellow-nosed in breeding d) Atlantic yellow-nosed in non-breeding Figure 3. Distribution of seabirds in ICCAT area. Color legend shows the utilization distribution (%) calculated from tracking data. Circle plot shows the bycatch CPUE calculated at each 5x5 grid square from observer data in Wandering albatross tracking data were provided for the purposes of this analysis with permission from Richard Phillips (British Antarctic Survey, UK), David Nicholls (Chisholm Institute, Australia), Deon Nel (WWF, South Africa), Henri Weimerskirch (CEB CNRS, France), and Rachael Alderman (DPIPWE, Australia) Atlantic yellow-nosed albatross tracking data were likewise provided with permission from Richard Cuthbert (Royal Society for the Protection of Birds, UK). 1793

11 a) Grey-headed albatrosses during breeding b) Grey-headed albatrosses during non-breeding c) White-chinned petrels during breeding d) White-chinned petrels during non-breeding Figure 4. Distribution of seabirds in ICCAT area. Color legend shows the utilization distribution (%) calculated from tracking data. Circle plot shows the bycatch CPUE calculated at each 5x5 grid square from observer data in Grey-headed albatross tracking data were provided for the purposes of this analysis with permission from Richard Phillips (British Antarctic Survey, UK), Deon Nel (WWF, South Africa), Graham Robertson (Australian Antarctic Division), Henri Weimerskirch (CEB CNRS, France), Javier Arata (Instituto Antártico Chileno), Rachael Alderman (DPIPWE, Australia), Rosemary Gales (DPIPWE, Australia). White-chinned petrel tracking data were provided for the purposes of this analysis with permission from Richard Phillips (British Antarctic Survey), Henri Weimerskirch (CEB CNRS, France), Leigh Torres (NIWA, New Zealand), and David Thompson (NIWA, New Zealand). 1794

12 Breeding Non-breeding Figure 5. Distribution of interaction (bird distribution x fishing effort) of all 7 species during breeding / non-breeding interaction.. The tracking data were provided for the purposes of this analysis with permission from Richard Phillips (British Antarctic Survey), Dave Watts (Australian Antarctic Division),David Nicholls (Bandicoot Recovery, Chisholm Institute, Australia), Deon Nel (WWF South Africa), Flavio Quintana (Wildlife Conservation Society, Argentina), Graham Robertson (Australian Antarctic Division), Henri Weimerskirch (CEB CNRS, France), Javier Arata (Instituto Antártico Chileno), Jose Pedro Granadeiro (CESAM, Museu Nacional de História Natural, Portugal), Nic Huin (Falklands Conservation), David Thompson(National Institute of Water & Atmospheric Research, New Zealand), Leigh Torres (National Institute of Water & Atmospheric Research, New Zealand), Rachael Alderman (DPUPWE, Australia), Richard Cuthbert(Royal Society for the Protection of Birds), Rosemary Gales (DPIPWE, Australia), Ross Waless (BirdLife South Africa), Samantha Petersen (WWF South Africa). 1795

13 Breeding Non-breeding Figure 6. Distribution of interaction (bird distribution x fishing effort) of black-browed albatrosses during breeding/non-breeding. Black-browed albatross tracking data were provided for the purposes of this analysis with permission from Richard Phillips (British Antarctic Survey), Dave Watts (Australian Antarctic Division), Flavio Quintana (Wildlife Conservation Society, Argentina), Graham Robertson (Australian Antarctic Division), Henri Weimerskirch (CEB CNRS, France), Javier Arata (Instituto Antártico Chileno), Jose Pedro Granadeiro (CESAM, Museu Nacional de História Natural, Portugal), Nic Huin (Falklands Conservation), Rachael Alderman (Australia), Rosemary Gales (DPIPWE, Australia), Samantha Petersen (WWF South Africa). Breeding Non-breeding Figure 7. Distribution of interaction (bird distribution x fishing effort) of sooty albatrosses during breeding/non-breeding. Sooty albatross tracking data were provided for the purposes of this analysis with permission from Henri Weimerskirch (CEB CNRS, France), Richard Cuthbert (Royal Society for the Protection of Birds, UK), Ross Wanless (BirdLife South Africa). 1796

14 Breeding Non-breeding Figure 8. Distribution of interaction (bird distribution x fishing effort) of Tristan albatrosses during breeding/non-breeding. Tristan albatross tracking data were provided with permission from Richard Cuthbert (Royal Society for the Protection of Birds, UK), and Ross Wanless (BirdLife South Africa). Breeding Non-breeding Figure 9. Distribution of interaction (bird distribution x fishing effort) of wandering albatrosses during breeding / non-breeding. Wandering albatross tracking data were provided for the purposes of this analysis with permission from Richard Phillips (British Antarctic Survey, UK), David Nicholls (Chisholm Institute, Australia), Deon Nel (WWF, South Africa), Henri Weimerskirch (CEB CNRS, France), and Rachael Alderman (DPIPWE, Australia). 1797

15 Breeding Non-breeding Figure 10. Distribution of interaction (bird distribution x fishing effort) of Atlantic yellow-nosed albatrosses during breeding/non-breeding. Atlantic yellow-nosed albatross tracking data were provided with permission from Richard Cuthbert (Royal Society for the Protection of Birds, UK). 1798

16 Breeding Non-breeding Figure 11. Distribution of interaction (bird distribution x fishing effort) of grey-headed albatrosses during breeding/non-breeding. Grey-headed albatross tracking data were provided for the purposes of this analysis with permission from Richard Phillips (British Antarctic Survey, UK), Deon Nel (WWF, South Africa), Graham Robertson (Australian Antarctic Division), Henri Weimerskirch (CEB CNRS, France), Javier Arata (Instituto Antártico Chileno), Rachael Alderman (DPIPWE, Australia), Rosemary Gales (DPIPWE, Australia). 1799

17 Breeding Non-breeding Figure 12. Distribution of interaction (bird distribution x fishing effort) of white-chinned petrels during breeding/ non-breeding. White-chinned petrel tracking data were provided for the purposes of this analysis with permission from Richard Phillips (British Antarctic Survey), Henri Weimerskirch (CEB CNRS, France), Leigh Torres (NIWA, New Zealand), and David Thompson (NIWA, New Zealand). 1800

18 Total seabirds Jan.-Mar. Apr.-Jun. Jul.-Sep. Oct.-Dec. + 0 Figure 13. CPUE distribution in the southern hemisphere of observer data of total seabirds used in this study by quarter of year during

19 Albatrosses Jan.-Mar. Apr.-Jun. Jul.-Sep. Oct.-Dec. + 0 Figure 14. CPUE distribution in the southern hemisphere of observer data of albatrosses used in this study by quarter of year during

20 Giant petrels Jan.-Mar. Apr.-Jun. Jul.-Sep. Oct.-Dec. + 0 Figure 15. CPUE distribution in the southern hemisphere of observer data of giant petrels used in this study by quarter of year during

21 Petrels Jan.-Mar. Apr.-Jun. Jul.-Sep. Oct.-Dec. + 0 Figure 16. CPUE distribution in the southern hemisphere of observer data of petrels used in this study by quarter of year during

22 Black-browed albatrosses Jan.-Mar. Apr.-Jun. Jul.-Sep. Oct.-Dec. + 0 Figure 17. CPUE distribution in the southern hemisphere of observer data of black-browed albatrosses used in this study by quarter of year during

23 Sooty albatrosses Jan.-Mar. Apr.-Jun. Jul.-Sep. Oct.-Dec. + 0 Figure 18. CPUE distribution in the southern hemisphere of observer data of sooty albatrosses used in this study by quarter of year during

24 Wandering albatrosses Jan.-Mar. Apr.-Jun. Jul.-Sep. Oct.-Dec. + 0 Figure 19. CPUE distribution in the southern hemisphere of observer data of wandering albatrosses used in this study by quarter of year during This data include bycatch of Tristan albatrosses. 1807

25 Yellow-nosed albatrosses Jan.-Mar. Apr.-Jun. Jul.-Sep. Oct.-Dec. + 0 Figure 20. CPUE distribution in the southern hemisphere of observer data of yellow-nosed albatrosses used in this study by quarter of year during

26 Grey-headed albatrosses Jan.-Mar. Apr.-Jun. Jul.-Sep. Oct.-Dec. + 0 Figure 21. CPUE distribution in the southern hemisphere of observer data of grey-headed albatrosses used in this study by quarter of year during

27 Grey petrels Jan.-Mar. Apr.-Jun. Jul.-Sep. Oct.-Dec. + 0 Figure 22. CPUE distribution in the southern hemisphere of observer data of grey petrels used in this study by quarter of year during

28 White-chinned petrels Jan.-Mar. Apr.-Jun. Jul.-Sep. Oct.-Dec. + 0 Figure 23. CPUE distribution in the southern hemisphere of observer data of white-chinned petrels used in this study by quarter of year during

29 Appendix 1 Tracking data provided for this study from the Global Procellariiform Tracking Database ( with permission from data owners. Sites may include a number of breeding colonies, of which not all may have been tracked. Sites with >1% global population for which tracking data are not available are also shown, to indicate major data gaps. Species Colony Number of tracks (breeding) Number of tracks (non-breeding) White-chinned Petrel Antipodes Islands Iles Crozet 16 0 South Georgia Auckland Is. 0 0 Falkland Is. / Islas Malvinas 0 0 Iles Kerguelen 0 0 Prince Edward Is. 0 0 Campbell Is. 0 0 Macquarie Is. 0 0 Wandering Albatross Iles Crozet Iles Kerguelen 11 0 Macquarie Island 4 4 Non-breeding, site unknown - 7 Prince Edward Islands 23 2 South Georgia Tristan Albatross Tristan da Cunha group unknown status Sooty Albatross Tristan da Cunha group unknown status Ile Amsterdam 0 5 Iles Crozet 26 0 Prince Edward Is. 0 0 Grey-headed Albatross Campbell Island 5 0 Chile 67 1 Macquarie Island 9 1 Prince Edward Islands 6 0 South Georgia unknown status Iles Crozet 0 0 Iles Kerguelen 0 0 Atlantic Yellow-nosed Tristan da Cunha group Black-browed Albatross Chile unknown status Falkland Islands / Islas Malvinas Heard & McDonald Islands 10 0 Iles Kerguelen 26 0 Macquarie Island 7 2 Non-breeding, site unknown - 8 South Georgia