Abstract:

Understanding key species and their roles and impacts within communities plays a vital role in protecting key species within communities. Species relationship indicators, i.e., betweenness centrality (BC), closeness centrality (CC), and degree centrality (DC) measured food web connectivity among a community of 49 species. Once indicators were calculated, a clustering analysis analyzed species influences on connectivity robustness among different species classes. In addition to clustering analysis, node removal was tested to measure connectivity robustness under removal modes. Based off of clustering results, species were divided into three food web classes: (1) Class I species have relatively higher BC, CC, and DC values. (2) Class II species have relatively lower BC, CC, and DC values while (3) Class III species are intermediate between Class I and Class II species. In our analyses, it appeared that Class I species are more important for food web processes and are therefore considered key species, contributing to our understanding of how to quantify key food web species. Besides, the connectivity robustness decreased along with the increase in removed nodes under the ascending order removal mode. In the descending order and random removal modes, the connectivity robustness slightly increased after a continuous decrease along with the increase in removed nodes. It indicates that removal modes significantly affect the connectivity robustness. Moreover, under the three removal methods, the significant change in the connectivity robustness corresponds to almost the same number of removed nodes. It suggests key species play an essential role in maintaining the connectivity robustness of the food web.

Key words: cluster analysis, complex network, ecological network, node degree, connectivity robustness