Title:Stability of complex communities: A perspective from discrete-time dynamics

Abstract:Understanding the stability of complex communities is a central focus in ecology, many important theoretical advancements have been made to identify drivers of ecological stability. However, previous results often rely on the continuous-time dynamics, assuming that species have overlapping generations. In contrast, numerous real-world communities consist of species with non-overlapping generations, whose quantitative behavior can only be precisely represented by discrete-time dynamics rather than continuous ones. Here, we develop a theoretical framework and propose a metric to quantify the stability of complex communities characterized by non-overlapping generations and diverse interaction types. In stark contrast to existing results for overlapping generations, we find that increasing self-regulation strength first stabilizes and then destabilizes complex communities. This pattern is further confirmed in both exploitative (E. aerogenes, P. aurantiaca, P. chlororaphis, P. citronellolis) and competitive (P. putida, P. veroni, S. marcescens) soil microbial communities. Moreover, we show that communities with diverse interaction types become the most stable, which is corroborated by empirical mouse microbial networks. Furthermore, we reveal that the prevalence of weak interactions can stabilize communities, which is consistent with findings from existing microbial experiments. Our analyses of complex communities with non-overlapping generations provide a more comprehensive understanding of ecological stability and informs practical strategies for ecological restoration and control.