台灣東部海域密毛龍蝦資源評估研究

Abstract

龍蝦為台灣沿海小型漁業之高經濟漁獲物種，然而有關其漁業及資源狀況的了解卻相當有限，有鑑於此，本研究以台灣東部海域密毛龍蝦為例，建構一套完整的量化模擬及資源評估理論架構，期對台灣的龍蝦資源有進一步的了解與掌握。成長模式是量化族群動態時最重的一環，本研究回顧目前所有量化甲殼類成長的模式，比較不同成長模式在不同物種的應用，並探討甲殼類成長與各種生物及非生物因子間的關係。 龍蝦漁業評估往往受資料及生活史過程之變異等未確定性所影響，本研究探討龍蝦的自然死亡率，並利用單位加入模式配合蒙地卡羅模擬法探討各種參數的未確定性對該龍蝦漁業評估結果的影響，結果發現單位加入模式中生物及漁業參數之未確性會影響生物參考點(F0.1及F40%)的估計。當漁業死亡率及生物參考點皆存在很高的未確定性下，其資源評估結果則存在很大的未確定。 為了詳細模擬龍蝦複雜的生活史，本研究建構個體模式來量化龍蝦的族群動態，進行台灣東部沿海密毛龍蝦資源評估，模式套適資料包括歷年龍蝦漁獲量及體長頻度資料。模式評估結果顯示，台灣東部沿海密毛龍蝦的產卵母群量為未開發程度的23%。根據貝氏後驗分布進行未來資源狀況之風險分析，結果顯示目前漁業利用率下，未來產卵親魚量低於未開發產卵母群量之40%的風險很低；然而以未開發產卵母群量之20%來看，則有較高的風險。此外，本研究進一步探討模式中各種生物及漁業參數對龍蝦族群動態及漁業的影響，結果建議體長限制比其他漁業管理策略對於保護龍蝦資源量更為有效。 環境變遷為影響龍蝦族群動態的重要因子，本研究利用所建構的個體模式探討氣候變遷對於生物參考點及漁業管理意涵的影響。結果證明了海水溫度變暖會提高龍蝦的生產力，並減少未來資源過度開的風險。然而，若海水溫度變暖伴隨較高的自然死亡率，則未來資源有較高的過度開風險。在不利的環境條件下，提高最小合法捕抓體長為有效漁業管理策略。在目前漁業資料缺乏的情況下，本研究建議利用所發展的個體模式來進行台灣海域龍蝦之資源評估，未來在模式中更明確地考慮環境因子和生物過程的關係，可改善資源評估結果的準確性。

Spiny lobsters are highly prized species for the small-scale fisheries in the coastal waters of Taiwan, however the information of the fisheries and the status of stocks are limited and unclear in the past. According to these concerns, the objective of this study is to develop a simulation testing and stock assessment framework for spiny lobster fisheries in Taiwan using the pronghorn spiny lobster (Panulirus peniciliatus) in the waters of eastern coast off Taiwan (Taitung) as an example. Because growth is one of the most important life history processes to study the population dynamics, I widely review the current approaches for modelling the growth of the crustaceans. A comparative study by adapting various methods were applied to different taxa of crustaceans. Lobsters’ fishery assessments are often subject to large uncertainty in input data and natural variability in life history, I evaluated the natural mortality of P.peniciliatus and used a Monte Carlo approach to incorporate life history parameters’ uncertainty into the estimation of the biological reference points (BRPs). The commonly used BRPs derived from the per recruit model (F0.1 and F40%) was suggested to be influenced by uncertainties associated with lobster life history and fishery parameters. A large uncertainty in the current fishing mortality rate and estimates of BRPs increased the uncertainty in determining the risk of overexploitation. To mimic the complex life history and fishing processes for P.peniciliatus, I further developed an individual-based model (IBM). The developed IBM was used to perform stock assessment, as the model was fitted to the catch and length-frequency of data collected from whole seller. The results indicated that the current spawning stock biomass remains at 23 % of its unfished level (SSB0). Decision analysis based on samples from a Bayesian posterior distribution indicated that there is a negligible risk of the stock dropping below 40% of SSB0 if fishing intensity remains at the current level. However, there is a high risk of overexploitation based on the reference level of 20% of SSB0. Using the developed IBM, I also examined the possible role of various key variable/parameters in characterizing the lobster population dynamics and lobster fishery. The result suggested that the size regulation is more effective compared to other technical measures. To include the environmental change into the population dynamics model, I applied the proposed IBM to evaluate potential impacts of increased ocean temperature on the estimation of mortality-based BRPs for fisheries management. A warming temperature would increase the stock productivity and consequently reduce the risk of overexploitation. However, there is likely a high risk of overexploitation in the long term if higher temperatures induce extra-high natural mortality. The evaluation of effectiveness of size regulations suggests that increasing minimum legal size is proposed as a good candidate measure to reduce the risk of overexploitation for pessimistically unfavorable environmental conditions. Finally, I recommended that using IBM to perform stock assessment in the current data-limited stage. An explicit incorporation of the relationships between environmental variables and biological processes can greatly improve the accuracy of stock assessment.