Abstract: Independent fishery surveys serve as a crucial pathway to ascertain the status of fishery resources within aquatic germplasm resource conservation areas. A well-designed sampling approach makes it possible to not only reduce the disturbance to fishery resources but also mitigate potential adverse effects on the ecosystem. We used fisheries data from single-vessel bottom trawl surveys conducted in the northern Beibu Gulf. The survey times including six voyages: spring (May) and autumn (September) of 2014, spring (April) and autumn (September) of 2018, autumn (November) of 2020, and spring (March) of 2021. We selected fish, cephalopods, shrimp, and crabs as our subjects of study. The resource density (kg/km²) of these groups was the target for our investigation, aiming to analyze the stability of the current station design over time. Additionally, we employed the ordinary kriging (OK) interpolation method to simulate the "true" distribution of resource densities for different biological communities. Based on this, the study established a simulation framework with six sampling designs: simple random sampling (D1), stratified random sampling by depth (D2), stratified random sampling by conservation area characteristics (D3), stratified random sampling by conservation area geographical boundaries (D4), stratified random sampling combining depth and geographical boundaries (D5), and stratified random sampling with a 30′×30′ spatial resolution (D6). We used the design effect (deff), relative estimation error (REE), and relative bias (RB) metrics to evaluate these designs, in an effort to test the accuracy and precision of different sampling designs on the assessment results. First, our research found that the current station design (D6) is not affected by temporal changes or the number of stations, demonstrating good stability in estimating the resource density of the four taxa in the area. When a comparable number of stations are deployed for the survey, the REE and deff values obtained using the D6 scheme are consistently smaller than those of D1 through D5. Although the RB values for cephalopods and crabs investigated using D6 may be overestimated (RB>8), D6 still exhibits better temporal stability than D1 through D5. Second, different station designs showed different trends in REE as the number of stations increased. The REE of D1 through D5 consistently declined, whereas D6 did not exhibit a clear, consistent trend of either decline or increase. Third, the survey accuracy of different station designs is closely related to the spatial distribution of the taxa. Since the distribution of some taxa does not consistently follow isobaths over time, D2 is deficient in temporal stability. Our study suggests that using D3 for cephalopod surveys should be approached with caution. The results show that when the station design is well-matched with the spatial distribution of various groups, survey accuracy can be significantly improved. We believe that the optimal station design varies slightly with both time and the number of stations. When the requirement is to deploy 9 stations, we advise implementing the D4 plan for conducting fishery resource surveys during the autumn voyage. For the spring season, no specific recommendation is made, as the optimal station design fluctuates among different species groups. However, when deploying either 12 or 15 stations, the D6 plan emerges as the preferred option for investigating the resource density of the four species groups in both spring and autumn, as its design is intended to maximize benefits. This study can provide a valuable reference for optimizing sampling designs in aquatic germplasm resource conservation areas.