【摘要】 传统的渔业资源评估方法需以翔实的调查和渔业数据为基础,而现有的大多数种类面临着着渔获量、基础生物学、有效捕捞努力量等数据缺失问题,因此并不适合采用数据需求较高的模型进行评估和管理。面临着渔业资源衰退的严峻形势和渔获量限额管理的迫切要求,基于有限数据的评估方法和渔获量相关的管理方案正被越来越多的国家采用。本研究以东海小黄鱼（Larimichthys polyactis）种群为例,根据渔获量、自然死亡、消减率、生物学参数、开捕体长等数据,采用54种有限数据评估方法,模拟3种捕捞动态,对小黄鱼进行管理策略评价和资源评估。结果显示,以相对产量（relative yield, RY）不低于50%、过度捕捞概率（probability of overfishing, POF）小于50%,生物量低于最大可持续生物量的10%(B<0.1B_MSY)的概率小于20%为风险控制水平,捕捞强度随机波动和增长情景下,分别有6个管理方案（management procedures, MPs）满足既定管理目标;"一般型"和"增长型"捕捞强度情景下, 14个MPs满足管理目标。权衡分析3种捕捞动态下的MPs, 50%FMSY基准法（FMSYref50）可作为小黄鱼渔业最佳的管理方案,POF介于5.46%～6.70%, B<0.5B_MSY概率介于15.66%～22.73%,长期获得的相对产量介于52%～100%;然而, FMSYref50确定的可接受生物学渔获量（acceptable biological catch, ABC）仅有1.08×10⁴ t,与当前产量相差较大。因此,考虑到降低捕捞强度为渔业管控的发展趋势,建议采用动态F比值法（DynF）为小黄鱼渔业管理方案,"下降型"捕捞强度情景下,POF为37.84%, B<0.5B_MSY概率为38.63%,长期获得的相对产量为84%, ABC为4.03×10⁴ t。根据敏感性分析,发现DynF评估的ABC对捕捞产量、资源丰度指数不敏感,而对自然死亡系数、最大可持续捕捞死亡系数与自然死亡系数比值（FMSY_M）和当前资源量均较为敏感,参数值增加会导致ABC增加,表明在开展渔业资源评估时需要着重提高这3种参数的准确性。更多还原

【Abstract】 The most established basis for estimating an Acceptable Biological Catch was by a conventional stock assessment, which typically used ?shery time series data to estimate current stock size and productivity. However,majority of ?sh stocks lack adequate description of catches, surveys, efforts, or information about life history characteristics to support a conventional stock assessment. Recent requirements to set scienti?cally-based catch limits, along with stock depletion and growing consumer demand for sustainably managed ?sheries, have created an emerging number of methods for estimating over?shing thresholds and setting catch limits for stocks with limited data. This research aimed to evaluate methods that determine an ABC as a basis for setting annual catch limits for small yellow croaker, Larimichthys polyactis. Using a management strategy evaluation approach, 54 established management procedures（MPs） for setting catch-limits in ?sheries with three fishing effort trend scenarios were compared. Performance was evaluated with respect to overfishing, biomass, and yield. According to the trade-offs between the expected relative yield, the probability of over?shing（POF）, and the probability of the biomass being below three different reference points in which the relative yield was not less than 50%, POF was less than 50% and the probability of B<0.1 B_MSY was less than 20%. Our results indicated that there were six MPs that met the established management target under both the generic fleet and increased fleet scenario simulations,while fourteen MPs met management targets under the decreased fishing mortality scenario simulation.FMSYref50 was considered to be the best MP for the yellow croaker fishery within three fishing mortality scenarios. Under the FMSYref50 MP, POF varied with a median of between 5.46% and 6.7%, the probability of B<0.5 B_MSY was between 15.66% and 22.73%, and the long-term relative yield varied from 52% to 100%. However,the ABC calculated under FMSYref50 was only 10800 tons, which would lead to a sharp decline in production compared with the current state of the fishery. The DynF MP was therefore suggested as the management strategy for the small yellow croaker fishery in consideration of the requirement of reducing numbers of fishing fleets.Under the decreased fishing mortality scenario simulation, the probability of overfishing was 37.84%, the probability of B<0.5 B_MSY was 38.63%, the long-term relative yield was 84% under the DynF MP, and the ABC was 40300 tons. The sensitivity analysis showed that the ABC allowed by the DynF MP was robust to the uncertainty of production data and abundance index, however, the ABC was sensitive to high imprecision in natural mortality rate, FMSY_M, and current biomass and provided more yield on average given increased observations. This indicated that the accuracy of parameters should be emphatically improved in conducting stock assessments with the DynF MP.更多还原