利用貝氏統計估計產卵群-加入群關係之陡度

Abstract

產卵群-加入群曲線決定加入群如何隨產卵群而變動,可以用來估計生物參考 點如FMSY,在漁業生物資源評估中為一重要數量分析模式。此模式之參數,陡度 (steepness)常用來重新參數化產卵群-加入群模式,亦為決定此曲線形狀之重要參數。 陡度為產卵群在其原始水平之20%所產生之加入群與原始水平加入群的比值,其 理論數值範圍介在0.2至1.0。較高的陡度值表示此生物系群在遭到開發後,擁有較 佳的回復力,且加入群的變動主要受到環境影響而非產卵群。過去進行太平洋黑 鮪(Thunnus orientalis)資源評估時,多以假設陡度值為1.0並以敏感度分析測驗。本 研究利用貝氏統計重新估算太平洋黑鮪的產卵群-加入群曲線之陡度值;以生產量 模式在合併與未合併漁業情況下估算系群之生物量,並設定具訊息與非訊息的事 前機率分布來估算事後機率分布。兩種不同的事前機率分布設定所產生的事後機 率分布差異不大(其事後分布之平均值分別為:訊息者為0.98,非訊息者為0.94), 顯示太平洋黑鮪確實具有高陡度值的潛力;即其加入群量受到產卵群密度的影響 較小。

To formulate the stock-recruitment (S-R) curve is one of the essential tasks in fishery stock assessment. Steepness is generally used to re-parameterize the S-R relationship thereby providing insight on resilience of a stock under exploitation. High steepness implies that a stock is relatively resilient. In this study, we use a Bayesian approach to re-estimate steepness of the Beverton-Holt S-R curve for the Pacific bluefin tuna (Thunnus orientalis) on the basis of the production model incorporating multiple fisheries. Previous steepness estimates (h ~ 1) for Pacific bluefin tuna seem too high to be plausible. Substantially, we evaluate the effects of using an uninformative prior vs. an informative prior based on information from other studies on posteriors of steepness. Our analysis shows small discrepancy between the two priors on their posteriors. The estimations of steepness (0.98 from vague-prior setting, and 0.94 from informative-prior setting) suggest that Pacific Bluefin tuna may be sensitive to variable environmental conditions.