Allocating observer sea days to Taiwanese distant water tuna longline fishing vessels
研究生: 林育宣 Yu-Xuan Lin
隨國際資源保育趨勢的發展，聯合國糧農組織及區域性漁業管理組織通過海鳥、海龜、鯊魚及鯨豚等重要混獲物種之保育措施，建議各國蒐集相關資料，並實施保育與管理措施。眾多資料蒐集方法中，觀察員計畫可提供混獲之詳實訊息，以供資源評估。但礙於經費等因素，並非所有漁船均皆可派遣觀察員登船紀錄，因此，本研究目的在於探討觀察員派遣方式，控制關鍵物種CPUE之變異係數，以提高資料可信度。
本研究以2004 ~ 2007年間我國168航次遠洋鮪延繩釣漁船觀察員資料為基礎，以目標魚種及混獲物種 (包含鯊魚、海鳥、海龜及鯨豚類)單位努力漁獲量 (CPUE)作為分層依據，將三大洋依據季節、作業區域分層，利用拔靴法 (bootstrap)估算每分層內各物種之單位努力漁獲量，以及對應不同觀察員涵蓋率下之物種CPUE變異係數。並依據混獲物種之分布，選擇各分層之重要關鍵物種，估算在關鍵物種CPUE之變異係數設定在30%所需之觀察員涵蓋率。進而以各分區關鍵物種CPUE變異係數為分派基準，結合總派遣天數以及區域性漁業管理組織要求的涵蓋率，估算最適涵蓋率。
結果顯示可將三大洋分成13區，其中南大西洋及西南印度洋區因分區內4季分層資訊齊全，資料筆數大於30且關鍵物種有顯著差異，予以季節分層。各分層關鍵物種變異係數控制在30%所需之觀察員涵蓋率最高者為南大西洋第1季 (關鍵物種為海鳥) 50%，最低者為南大西洋第2季、中大西洋、西北太平洋、西南太平洋及西南印度洋第1季皆為1%。以2009年觀察員62人換算全年可派遣最高執勤天數9548天進行分配，故建議全區關鍵物種CPUE變異係數控制於25%，大西洋平均涵蓋率為7.33%，太平洋為15.70%，印度洋為11.81%以上，區域分層方面，觀察員涵蓋率最高者為中西太平洋區之55%，最低者為東南印度洋區之4.58%。
總結而言，並非所有分區之觀察員派遣涵蓋率皆需達5%，才能使資料變異係數控制於30%，在中大西洋區、南大西洋區的第1、3季及西北及西南太平洋等區域可以降低觀察員努力量，移至其他區域，如南大西洋第1季、中西太平洋及東北印度洋，故未來在執行觀察員分派時應做整體考量，兼顧資料可信度、人力負荷以及國際組織要求。
Following the international trend on marine resources conservation, Food and Agriculture Organization of the United Nations and many Regional Fisheries Management Organizations adopted measures for conservation of important incidental catch species, include seabirds, sea turtles, sharks and marine mammals. Those organizations recommend related countries shall collect data and implement conservation and management measures. Among those data collection method, observer program could provide detail incidental catch information for evaluate the fisheries impacts on those bycatch species. However, not all fishing vessels could deploy observers because of budget limitation. The objective of this research is to discuss the observer coverage by control the coefficient of variance (CV) of catch per unit effort (CPUE), to improve the precision.
One hundred and sixty eight trips of Taiwanese distant water longline vessels observer trips data from 2004 to 2007 were used for analysis. The catch per unit effort (CPUE) of target species and bycatch species (including sharks, seabirds, seaturtles, and cetaceans) are used for spatial and temporal stratification. Bootstrap is used to estimate the CPUE and CV for each stratum under different observer coverage from 1% to 100%. The key conservation species are selected in accordance with their distribution. The observer coverage rate of key species under the CV =30% are considered as the minimum requirement for observer coverage requirement. In addition, the request from Regional fisheries management organizations, and total observers day are considered as well.
The results showed thirteen sub-regions are separated. Among these sub-regions, south Atlantic Ocean and southwest Indian Ocean are further separated by seasons because the data are more than 30 in each stratum and the CPUE of key species are significant difference for seasons. Among these strata, the coverage rate ranged from 50% (first season of south Atlantic Ocean for seabirds) to 1% (second season of south Atlantic Ocean, central Atlantic Ocean, northwest Pacific Ocean, southwest Pacific Ocean and first season of southwest Indian Ocean). The total observer days is 9548 days based on the number of observers was 62. We could control the CV at 25% for three oceans; which the coverage rate was 7.33% for Atlantic Ocean, 15.7% for Pacific Ocean, and 11.81% for Indian Ocean. Among these sub-regions, the observer coverage rate was highest in central western Pacific Ocean (55%), and the lowest in the southeast Indian Ocean (4.58%).
In conclusion, it is not necessary for all areas to deploy 5% observer to ensure the CV less than 30%. The central Atlantic Ocean, second and third season of south Atlantic Ocean, northwest and southwest Pacific Ocean could have less observer coverage. The additional observer could reallocate to other area, such as first season of south Atlantic Ocean, central western Pacific Ocean and northeast Indian Ocean. In the future, when deploying the observer, shall considered the data precision, total capacity of observer program and request from regional fisheries management organizations.
