The commercial fishery for subtidal Pacific geoduck clams (Panopea generosa) in Washington State, USA, is substantial both by the amount of biomass extracted (2 million kg in 2019) and by the economic value (US $50 million annual). Management for this fishery, which began in 1970, is challenged by this species’ long lives, cryptic behavior, and recruitment variability. Current management, including a 2.7 % annual harvest rate, is based on an equilibrium yield model most sensitive to a natural mortality parameter estimated using sampled age distributions. We collected a large sample for aging from four new sites in Puget Sound to update the yield model and explore its applicability to individual management regions. Based on the new age distributions, natural mortality varies among regions, but is similar overall (0.0231 yr⁻¹) to the value used in original formulation of the yield model (0.0226 yr⁻¹), indicating that a change to the harvest rate is not needed if following the outputs of this model. However, concerns remain about the sustainability of the current harvest rate, particularly the time for harvested tracts (delineated harvest areas ranging in size from 0.06 to 1.45 km²) to return to their preharvest density. We calculated tract recovery rates from serial post-harvest density surveys, compared them to the expected recovery rate derived from the age-based equilibrium model, and investigated possible mechanisms behind fast or slow tract recovery. Using multiple scuba surveys spaced over four decades from 38 tracts in the South Puget Sound, we estimate that tracts recover on average at 0.03 geoducks per square meter per year. The projected average time to recovery is 55 years, compared to a projection of 39 years when the yield model was developed. There is some support for the hypothesis that tract recovery rates have slowed in recent decades. Although there were strong spatial patterns, with higher recovery in central basins compared to within inlets, there was only weak support for the hypothesized mechanisms behind tract recovery such as mean current speed. Our inability to clearly relate recovery to physical variables such as habitat, paired with extreme interannual variability in recruitment suggested by the age distributions, supports the hypothesis that the strength, timing, and spatial extent of episodic recruitment events is the primary driver of tract recovery. Currently, harvestable biomass estimation and fishery management assume that all previously harvested tracts will eventually recover to preharvest densities. However, in the South Puget Sound, only 40 % of tracts are likely to recover on a 50-year time frame relevant to management. Despite our validation of the key parameter used in the yield model, we advocate switching geoduck management to a tract-recovery-based strategy in which harvest is paced to known tract recovery, and biomass estimation is restricted to only those areas likely to hold renewable resource. This study highlights the importance of evaluating recovery patterns of stocks once data over appropriate timescales are available, and comparing to the management model to determine if fishery goals are being met.

潮下太平洋象拔蚌 ( Panopea generosa)的商业渔业) 在美国华盛顿州，无论是从提取的生物质数量（2019 年为 200 万公斤）还是经济价值（每年 5000 万美元）方面都是巨大的。该渔业的管理始于 1970 年，受到该物种寿命长、行为隐秘和招募变异性的挑战。目前的管理，包括 2.7% 的年收成率，是基于对使用采样年龄分布估计的自然死亡率参数最敏感的平衡产量模型。我们从普吉特海湾的四个新地点收集了一个大样本进行老化，以更新产量模型并探索其对各个管理区域的适用性。根据新的年龄分布，自然死亡率因地区而异，但总体相似（0.0231 yr ^-1) 到产量模型原始公式中使用的值 (0.0226 yr ^-1 )，表明如果遵循该模型的输出，则不需要改变收获率。然而，人们仍然担心当前采伐率的可持续性，特别是采伐面积（划定的采伐面积范围为 0.06 至 1.45 平方公里²) 恢复到收获前的密度。我们从系列收获后密度调查中计算了道恢复率，将它们与从基于年龄的平衡模型得出的预期恢复率进行了比较，并研究了快速或缓慢道恢复背后的可能机制。使用南普吉特海湾（South Puget Sound）的38个水域进行了长达40年的多次水肺调查，我们估计这些水域平均每年每平方米可恢复0.03只象拔蚌。预计平均恢复时间为 55 年，而在开发产量模型时预计为 39 年。近几十年来，管道恢复率已经放缓的假设得到了一些支持。尽管存在强烈的空间格局，与入口内相比，中央盆地的采收率更高，对于道恢复背后的假设机制，例如平均当前速度，只有微弱的支持。我们无法清楚地将恢复与栖息地等物理变量联系起来，再加上年龄分布所暗示的招募的极端年际变化，支持了这样的假设，即偶发招募事件的强度、时间和空间范围是道恢复的主要驱动因素。目前，可收获生物量估算和渔业管理假设所有先前收获的区域最终将恢复到收获前的密度。然而，在南普吉特海湾，在与管理相关的 50 年时间框架内，只有 40% 的土地可能会恢复。尽管我们验证了产量模型中使用的关键参数，我们主张将象拔蚌的管理转变为基于地块恢复的策略，在该策略中，收获按照已知的地块恢复速度进行，并且生物量估计仅限于那些可能拥有可再生资源的区域。这项研究强调了在获得适当时间范围内的数据后评估种群恢复模式的重要性，并与管理模型进行比较以确定是否达到渔业目标。