Using network analysis to identify keystone species in the food web of Haizhou Bay, China

Abstract

Keystone species play critical roles in ecological communities and ecosystem functions by interacting directly or indirectly with other species. Studies on keystone species will contribute to the understanding of community stability and mechanisms underlying ecosystem degradation. Based on a survey in Haizhou Bay and adjacent waters in 2011, the food web was constructed according to the predator–prey relationships of species. The binary network was analysed to calculate 11 network indices to identify keystone species, and removal analysis was conducted to examine the effects of the extinction of keystone species on the food web. Kendall rank correlation analysis showed that there were significant correlations between most pairs of the 11 network indices (P<0.05). According to the results of principal component analysis of all these indices, several species, namely Leptochela gracilis, Loligo sp., Larimichthys polyactis, Alpheus japonicus and Oratosquilla oratoria, were identified as keystone species in the Haizhou Bay food web. Removal analysis suggested that the absence of keystone species may have considerable effects on the complexity and stability of the food web in Haizhou Bay. It is suggested that priority is given to the protection of keystone species in the marine ecosystem.