Ecosystems can exhibit rich equilibrium behaviors, including alternative stable states (ASS), for which ecosystems dramatically differ in structure and functions. Therefore, it is critical to understand system equilibrium behavior. Positive feedback has been recognized as a critical process to induce ASS. However, it is largely unclear whether and to which extent increasing number of positive feedbacks will be easier to produce ASS. Using a generalized Lotka-Volterra model and two functional forms of positive feedback, this study found that positive feedback was largely necessary to produce ASS, and more positive feedbacks generally more likely produced ASS. However, above results depended on the functional form of positive feedback and interaction types. This work can help predict ecosystem dynamics under global change and intensified anthropogenic activities, both which may activate previously “dormant” positive feedbacks and make this issue more visible and important.

Ecosystems may exhibit transient (the period toward stable state) and alternative stable states, where positive feedback(s) generally plays important roles. However, the outline between transient and alternative stable states may blur when involving multiple positive feedbacks, which may shift dominance along environmental gradient. Therefore, a common theoretical framework considering multiple positive feedbacks is needed. Specifically, this study employed a mechanic model to explain a long transient state of a dryland ecosystem after restoration under effects of two positive feedbacks. We found that the long transient state might be a temporal visit of a hidden alternative stable state, maintained by the previously on-acting positive feedback and then pulled out by the upcoming feedback due to dominance shifting. This work proposes a novel perspective to study system dynamics – switch of positive feedbacks, and discovers a new type of transient state – temporal visit of hidden alternative stable states, which will benefit ecosystem interventions in practical.