Trophic dynamics is one of the major regulators of fishery production in marine ecosystems, which is important for the implementation of ecosystem-based fisheries management. Based on data collected form bottom trawl surveys in Haizhou Bay and adjacent waters during autumn of 2011 and 2018, Delta-GAMMs (Delta-generalized additive mixed models) were constructed to evaluate the effects of biotic and abiotic factors on the predation of five key prey species (including Leptochela gracilis, Alpheus japonicus, Loligo spp., Larimichthys polyactis, and Oratosquilla oratoria) in the Haizhou Bay. Percent frequency of occurrence and predation pressure index were used to identify their major predators. Variance inflation factor and full subsets regression were analyzed to quantify the degree of multicollinearity between these factors. The results showed that the occurrence frequency of keystone prey species in the stomach of predators ranged from 8.5% to 42.2%, and the weight percentage ranged from 4.2% to 40.9%. The average deviance explanation rate of the "binomial" model was 16.1%, and the average deviance explanation rate of the "positive" model was 23.8%. Body length of predator, predator population density, and sea bottom temperature were important factors influencing prey-predator trophic interactions. Predator length was the most important factor, with feeding probability and weight percentage of keystone prey species all increasing with body length of predator. Feeding probability and weight percentage of key prey species decreased with predator population density. The effects of environmental factors such as sea bottom temperature, depth of water, latitude, and sea bottom salinity showed different trends, depending on the prey-predator assemblage. This study showed that the Delta-GAMMs was an effective method to explore the trophic interactions between prey and predators in marine ecosystems, and could provide a theoretical basis for the conservation and sustainable use of fisheries resources.
海洋生态系统的营养动力学是决定渔业生产力的主要因素,也是实施基于生态系统的渔业管理的重要基础。为了评估生物因子和非生物因子对海州湾5种关键饵料生物被摄食的影响,本研究基于2011和2018年秋季在海州湾及其邻近海域进行的渔业资源底拖网调查和胃含物分析数据,构建了Delta-GAMMs模型,选择细螯虾、日本鼓虾、枪乌贼、小黄鱼和口虾蛄5种关键饵料生物作为研究对象,基于出现频率百分比和捕食压力指数筛选其主要捕食者,基于方差膨胀因子和全子集回归筛选影响5种关键饵料生物被摄食的关键因子。结果表明: 5种关键饵料生物在捕食者食物中的出现频率范围为8.5%～42.2%,重量百分比范围为4.2%～40.9%。Delta-GAMMs二项模型和正态模型的平均偏差解释率分别为16.1%和23.8%。捕食者体长、捕食者种群密度和底层水温是影响饵料生物-捕食者相互作用的重要因素。其中,捕食者体长是最重要的影响因素,关键饵料生物的被摄食率和被摄食量均随捕食者体长的增加呈上升趋势,关键饵料生物的被摄食率和被摄食量随着捕食者种群密度的增加而下降。此外,底层水温、水深、纬度和底层盐度等环境因素对种间营养关系的影响,因饵料生物-捕食者组合的不同而呈现不同的趋势。研究表明,Delta-GAMMs模型是探究海洋生态系统饵料生物与捕食者相互作用的有效方法,能够为渔业资源养护和可持续利用提供理论依据。.
Keywords: Haizhou Bay; generalized additive mixed models; keystone prey species; trophic interaction.