西北太平洋秋刀魚族群動態與資源評估

Abstract

秋刀魚（Cololabis saira）為西北太平洋海域重要之國際漁業資源，其短壽命的生活史特性使族群時空動態易受環境變遷影響。近年來漁獲量持續下降，本研究建構完整之資源評估理論架構，針對秋刀魚族群動態進行方法學改善，以期對此資源有更全面之掌握。本研究從四個面向探討影響秋刀魚資源評估之關鍵要素：首先，運用演化生態學和生殖生物學之數值模擬分析，評估族群加入量（recruitment）之韌性（resilience），此方法解決漁業資源評估常因缺乏可靠且長期的產卵親魚量與入添量數據，而導致族群韌性估計存在高度不確定性之挑戰。結果顯示陡度（steepness）中位數為0.82（80%可能範圍：0.59 − 0.93），代表即使族群產卵親魚量降至未開發時之20%，仍可維持相對高比例的加入量。敏感度分析指出，陡度受幼生死亡率、平均體重、成長和性成熟體長等因素影響最顯著，而不利的環境條件可能透過影響這些參數而降低族群韌性。其次，應用廣義線性混合模型（Generalized Linear Mixed Models，GLMMs）評估不同空間處理方法對單位努力漁獲量（catch-per-unit-effort，CPUE）標準化之影響。研究發現豐度主要受時空效應影響，可能源於漁撈活動或魚群分布之時空變動。在實際漁業資料和模擬研究中，向量自回歸時空模型（Vector Autoregressive Spatio-Temporal，VAST）均優於傳統空間分區方法。當遇優先採樣情況，空間集群分析法中之Spatial1（魚類密度權重低）可作為VAST之替代方案。第三，運用VAST評估秋刀魚之時空動態。結果顯示族群分布重心在2013年後明顯由日本沿岸向公海東移，2017年更出現豐度顯著下降及進一步東移現象。研究發現此分布變化無法單由海洋環境因子或氣候變異指數解釋，推測可能與種間競爭、公海漁獲壓力增加或複雜海洋動態等因素相關。最後，開發秋刀魚年齡結構模型，並考慮諸多未確定性來源與加入量的過程誤差（process error）進行未來族群量預測。評估結果顯示近年產卵親魚量降至歷史低位，低於最大可持續生產量水準（SSBMSY）。儘管目前漁獲死亡率仍低於FMSY，預測結果顯示可能需要更保守的漁獲狀況以利資源恢復。其中，採用0.5FMSY情境顯示最高機率維持資源量於SSBMSY以上。本研究深化了對秋刀魚族群動態的理解，並改進了資源評估方法。研究成果強調在發展有效漁業管理策略時，應充分考量環境變異性、空間分布型態及參數不確定性之重要性。研究成果除可為西北太平洋秋刀魚之永續經營提供科學基礎外，所建立之方法論架構亦可作為其他類似漁業資源評估改進之參考。

Pacific saury (Cololabis saira) is an important international fishery resource in the Northwestern Pacific Ocean. Its short-lived life history characteristics make its population dynamics particularly susceptible to environmental changes. In response to recent declining catches, this study developed a comprehensive stock assessment framework, advancing methodological approaches to better understand Pacific saury population dynamics. This study systematically investigated four critical aspects of Pacific saury stock assessment. First, this study employed numerical simulation analyses integrating evolutionary ecology and reproductive biology to evaluate population recruitment resilience. This approach addressed a fundamental challenge in fisheries assessment: the high uncertainty in resilience estimation due to limited long-term data on spawning stock biomass and recruitment. Results suggested a median steepness of 0.82 (80% probable range 0.59, 0.93), indicating substantial recruitment capacity even when spawning biomass decreases to 20% of its unexploited level. Sensitivity analyses identified survival rate of early-life stages, mean body weight, growth, and length at maturity as key determinants of steepness, with unfavorable environmental conditions potentially compromising population resilience through these parameters. Second, this study evaluated the impact of spatial treatments on catch-per-unit-effort (CPUE) standardization using Generalized Linear Mixed Models (GLMMs). The analysis demonstrated that abundance was predominantly influenced by spatiotemporal effects, likely reflecting temporal shifts in fishing activities and fish distribution patterns. The Vector Autoregressive Spatio-Temporal (VAST) model consistently outperformed traditional methods in both empirical and simulation studies. For scenarios involving preferential sampling, the Spatial1 approach (incorporating low fish density weighting) emerged as a viable alternative to VAST. Third, the VAST analysis of spatiotemporal dynamics indicated a significant eastward shift in Pacific saury’s distribution from Japanese coastal waters to the high seas after 2013, followed by further abundance declines and eastward movement in 2017. Notably, these distributional changes could not be adequately explained by oceanic environmental factors or climate variability indices, suggesting potential influences from interspecific competition, increased high-seas fishing pressure, or complex oceanographic processes. Finally, this study developed an age-structured model incorporating multiple uncertainty sources and recruitment process error for forecast projections. The assessment indicated a historical low in recent spawning stock biomass, falling below SSBMSY. Despite current fishing mortality remaining below FMSY, projections suggested the need for more conservative harvest rates to facilitate stock recovery, with the 0.5FMSY scenario showing the highest probability of maintaining stock above SSBMSY. This research enhances the understanding of Pacific saury population dynamics while advancing stock assessment and management methodologies. The findings highlighted the importance of integrating environmental variability, spatial distribution patterns, and parameter uncertainty into effective fisheries management strategies. Beyond its application to Pacific saury, the established methodological framework provides valuable insights for improving stock assessments of similar fisheries worldwide.