Size selectivity of T90 mesh codends in stownet for Coilia mystus and Metapenaeus joyneri in the South China Sea

Stownet fishery is one of the most important component in coastal waters of China, characterized by simple gear structure, low energy consumption and stable economic benefits. In the South China Sea, two-stick stownet are widely used to harvest small fish and crustaceans, especially Coilia mystus and Metapenaeus joyneri. However, traditional stownet usually employ small mesh sizes and diamond-shaped meshes, resulting in poor size selectivity and high bycatch rates of juveniles, which pose serious threats to sustainable exploitation of coastal fishery resources. Codend structure, particularly mesh size and mesh orientation, plays a key role in determining the geometric configuration, hydrodynamic performance and selective properties of stownet. Compared with conventional diamond meshes, T90 meshes (diamond meshes rotated by 90°) have been proven to maintain more stable opening geometry and better selectivity in trawl fisheries, but their effectiveness in passive stownet systems remains insufficiently quantified. The objective of this study was to quantify the size selectivity of T90 mesh codends with different mesh sizes used in two-stick stownet fisheries in the South China Sea, and to evaluate their effects on the selective properties for C. mystus and M. joyneri, providing a scientific basis for codend design optimization and mesh size regulation. Field selectivity experiments were conducted using the covered method with T90 mesh codends of three mesh sizes (30, 35, and 40 mm). Catch data from the codend and the covered net were collected and sorted by size. Logistic and Richards models were applied to fit the size selectivity curves for C. mystus and M. joyneri, and selectivity parameters, including the 50% selection length (L₅₀), were estimated using the maximum likelihood method. The relationship between L₅₀ and mesh size was further analyzed to estimate the minimum appropriate mesh size for each target species. The results showed a clear increase in size selectivity with increasing mesh size. For C. mystus, the estimated L₅₀ values were 130.88 mm, 134.04 mm, and 142.72 mm for the 30, 35, and 40 mm T90 mesh codends, respectively. For M. joyneri, the corresponding L₅₀ values were 70.75 mm, 79.93 mm, and 90.10 mm. In both species, L₅₀ increased consistently with mesh size, indicating enhanced release of smaller individuals. Based on the linear relationship between L₅₀ and mesh size, the minimum appropriate mesh sizes were estimated to be 28.20 mm for C. mystus and 29.51 mm for M. joyneri. Considering the body shape characteristics of the target species and the selectivity results, it is recommended that the minimum mesh size of T90 codends used in two-stick stownet fisheries in the South China Sea should not be less than 30 mm. Compared with traditional diamond meshes, T90 mesh codends can maintain higher opening stability and significantly improve the release efficiency of juvenile fish and shrimp. These findings demonstrate that the adoption of T90 mesh structures represents an effective technical measure to improve stownet selectivity and contributes to the sustainable management of coastal fishery resources in the South China Sea.