Community ecologists have made great advances in understanding how natural communities can be both diverse and stable by studying communities as interaction networks. However, focus has been on interaction networks aggregated over time, neglecting the consequences of the seasonal organization of interactions, henceforth seasonal structure, for community stability. Here, we extended previous theoretical findings on the topic in two ways: (i) by integrating empirical seasonal structure of 11 plant-hummingbird communities into dynamic models, and (ii) by tackling multiple facets of network stability together. We show that, in a competition context, seasonal structure enhances community stability by allowing diverse and resilient communities while preserving their robustness to species extinctions. The positive effects of empirical seasonal structure on network stability vanished when using randomized seasonal structures, suggesting that eco-evolutionary dynamics produce stabilizing seasonal structures. We also show that the effects of seasonal structure on community stability are mainly mediated by changes in network structure and productivity, suggesting that the seasonal structure of a community is an important and yet neglected aspect in the diversity-stability and diversity-productivity debates.

We used data from 11 independent sites in the tropical forests of Ecuador (Table 1, Fig. S1) in which interactions among flowering plants and hummingbirds were recorded along transects by using camera traps, as described in Graham & Weinstein (2018).

Graham CH, Weinstein BG. Towards a predictive model of species interaction beta diversity. Ecol Lett. 2018;21(9):1299–310.