It is important in fisheries management to evaluate the effects of environmental factors on changes in the abundance of species, particularly those inhabiting estuaries. The Yangtze River estuary is the largest in the western Pacific Ocean and serves as a spawning, feeding and nursery ground to species of economic and ecological importance, such as Collichthys lucidus and Lophiogobius ocellicauda. By using 3 years of environmental variables and data on the abundance of C. lucidus and L. ocellicauda gathered seasonally through trawl surveys, we compared four generalised additive models (GAMs), each comprising different spline terms, to analyse the influence of the measured variables on the abundance of both species. Deviance explained, Akaike information criterion and generalised cross-validation were used to select the optimal GAM after evaluating the fit and predictive performance of the models. Generalised additive model (GAM) with spline ‘te’ was the optimal model, and predicted that the abundance of both species was influenced by season and by variables temperature, salinity and chlorophyll-a concentration. For C. lucidus, abundance increased during spring and summer, and, for L. ocellicauda, it was higher during winter. Given the socioeconomic importance of both species, we contend that determining the drivers causing abundance fluctuations of estuarine species can support the putting in place of robust monitoring and assessment plans for such fisheries.

在渔业管理中，评估环境因素对物种特别是河口物种丰富度变化的影响非常重要。长江口是西太平洋最大的河口，是诸如Collichthys lucidus和Lophiogobius ocellicauda等具有经济和生态重要性的物种的产卵，觅食和育苗场。通过使用3年的环境变量和有关透明念珠菌和L. ocellicauda的丰度数据通过拖网调查季节性收集的数据，我们比较了四个广义加性模型（GAM），每个模型都包含不同的样条项，以分析测量变量对两个物种的丰度的影响。Deviance解释说，在评估模型的拟合和预测性能之后，使用Akaike信息准则和广义交叉验证来选择最佳GAM。样条为“ te ”的广义加性模型（GAM）是最佳模型，并预测两种物种的丰度均受季节以及温度，盐度和叶绿素a浓度变量的影响。对于C. lucidus，在春季和夏季，丰度增加，对于L. ocellicauda，，冬季时较高。考虑到这两种物种的社会经济重要性，我们认为确定导致河口物种大量波动的驱动因素可以支持针对此类渔业建立强有力的监测和评估计划。