为了解东莞生态园湿地水域群落结构特征，探讨2种类型刺网 (沉网和浮网)及不同网目网片渔获物组成的差异性，实验利用多网目 (1.0、1.25、1.6、2.0、2.5、3.1、3.9、4.8、5.8、7.0、8.6和11.0 cm)单层刺网对该湿地的鱼类进行了调查。研究期间共采集到13种鱼类，相对重要性指数显示，海南似鱎和莫桑比克罗非鱼为优势种，前者密度最高 (占84.8%)，后者生物量居多 (40.2%)。渔获物物种组成沉浮网间差异显著，但单位努力捕捞数量 (NPUE)、单位努力捕获重量 (BPUE)及物种数与沉浮网间均无显著差异；海南似鱎的NPUE在2种类型刺网间无显著差异，但莫桑比克罗非鱼和尼罗罗非鱼的NPUE浮网显著低于沉网。12个网目网片的渔获物物种聚成2类，1.0~2.0 cm网目网片渔获物聚为1类，海南似鱎为指示种；2.5~11.0 cm片的聚为另1类，莫桑比克罗非鱼和尼罗罗非鱼为指示种。渔获物大小结构聚为2类，1.0~3.9 cm网目网片渔获物聚为1类，4.8~11.0 cm网目网片的聚为另1类。随着网目大小增加，渔获物平均体长线性递增，NPUE呈幂函数式下降。研究表明，为对内陆水域鱼类群落特征进行更加科学地估计，刺网需要增加1~2 cm网目网片，且需要同时放置沉网和浮网进行鱼类调查。
Gillnet is extensively used as fishing tool throughout the world. However, it has high species- and size-selectivity for fishes because of net mesh size. The selectivity caused by net mesh size will inevitably reduce comparability on attributes of fish assemblages sampled with non-standard gillnets. Therefore, the standard of fish sampling with multi-mesh gillnets has been universally established and applied in North America and Europe since more than ten years ago. The standard will enhance the comparability across large space and long time scale. Unfortunately, there is so far no analogous standard in China. In this study, using two types of multi-mesh gillnets (benthic and floating) with 12 mesh sizes of 1.0, 1.25, 1.6, 2.0, 2.5, 3.1, 3.9, 4.8, 5.8, 7.0, 8.6 and 11.0 cm, we sampled fishes from June to December, 2019 in Dongguan ecological park wetland, an important component of the Dongguan National Urban Wetland Park wetland. The differences in attributes of catch community between two type of gillnets and among different nets mesh sizes were explored. We hope that this study will help to establish Chinese standard of fish sampling with multi-gillnets. Meanwhile, we revealed the structural characteristics of catch community in the wetland to provide scientific basis for its management. The results showed that a total of 13 fish species were collected, and Toxabramis houdemeri and Oreochromis mossambica were the dominant species in terms of relative importance index. The former species accounted for 84.8 percent of the abundance of the whole catch community and the latter one contributed 40.2 percent to its biomass, respectively. There was a significant difference between benthic and floating gillnets in catch species composition, but there were no significant difference in number per unit effort (NPUE), biomass per unit effort (BPUE) and species number; floating gillnet had significantly lower NPUE of two tilapia species (O. mossambica and Oreochromis nilotica) than benthic gillnet, but T. houdemeri did not significantly differ in NPUE between the two type of gillnets. Two species clusters of catches sampled by nets with 12 different mesh sizes were obtained using cluster analysis (CA) and non-metric multidimensional scaling (NMDS); one species cluster was composed of catches collected by nets with mesh size of 1.0-2.0 cm with indicator species of T. houdemeri, and another species cluster consisted of catches sampled by nets with mesh size of 2.5-11.0 cm with indicator species of O. mossambica and O. nilotica. Meanwhile, two size clusters of all catches were also obtained using CA and NMDS; one size cluster was made up of catches sampled with 1.0-3.9 cm mesh-size nets and another size cluster comprised catches collected with 4.8-11.0 cm mesh-size nets. With increasing mesh size, the average body length of catch increased linearly, while NPUE decreased by power function. To more scientifically estimate the fish community structure in inland waters, adding nets with mesh size of 1 to 2 cm to the multi-mesh gillnet and combining use of benthic and floating gillnets would be recommended in the future.