Indirect effects among biodiversity loss of mutualistic ecosystems

Drastic reduction in biodiversity has been a severe threat to ecosystems, which is exacerbated when losing few species leads to disastrous and even irreparable consequences. Therefore, revealing the mechanism underlining biodiversity loss is of uttermost importance. In this study, we show that abundant indirect interactions among mutualistic ecosystems are critical in determining species' status. Combining topological and ecological characteristics, we propose an indicator derived from a dynamic model to identify keystone species and quantify their influence, which outperforms widely-used indicators like degree in realistic and simulated networks. Furthermore, we demonstrate that networks with high modularity, heterogeneity, biodiversity, and less intimate interactions tend to have larger indirect effects, which are more amenable in predicting decline of biodiversity with the proposed indicator. These findings shed some light onto the influence of apposite biodiversities, paving the way from complex network theory to ecosystem protection and restoration.