Role of species associations and environmental variables in structuring the coastal fish community: A case study in Leizhou Bay, northern South China Sea

The objective of this study was to understand how the biological and environmental factors affect fish community structures. Using the data of fish communities in Leizhou Bay waters, northern South China Sea, this study built an ecological network based on the model niche overlap between species. The effects of biological and environmental factors on the fish community structure complexity were analysed using principal component analysis (PCA) and generalised linear model (GLM), of which the complexity was represented with a global weighted clustering coefficient (mCw) of the network. Results showed that the fish community structure was highly complex with significant seasonal variations (mCw = 0.8211 +/- 0.081, p<0.01) and it was dominated by the strong interspecific associations (mCw>C). A large quantity of species associations decreased the chances of the cascading species extinction in terms of the magnitude of direct (k(i)) and indirect (tris) species associations. The direct species associations (k(i) : Estimate = -4.23, p<0.01) and the indirect species associations (iris: Estimate =3.12, p<0.01) were the primary factors, accounting for 89.92% of the variations of the first principal components and then the three factors (the number of species, water depth and the positions of sampling sites) contributed to 83.1% variations of the second principal components. In all, the species associations directly affected the fish community in Leizhou Bay, which were influenced by the species abundance under the driving of environmental factors.